View
4
Download
0
Category
Preview:
Citation preview
GENDER AND PERSONALITY DIFFERENCES IN COGNITIVE TASKS:
A PERFORMANCE AND MENTAL WORKLOAD STUDY
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
BETÜL BATUN
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR
THE DEGREE OF MASTER OF SCIENCE
IN
INDUSTRIAL ENGINEERING
JUNE 2015
Approval of the thesis:
GENDER AND PERSONALITY DIFFERENCES IN
COGNITIVE TASKS: A PERFORMANCE AND MENTAL
WORKLOAD STUDY
submitted by BETÜL BATUN in partial fulfillment of the requirements for
the degree of Master of Science in Industrial Engineering Department, Middle
East Technical University by,
Prof. Dr. Gülbin Dural Ünver
Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Murat Köksalan Head of Department, Industrial Engineering
Prof. Dr. Canan Çilingir Supervisor, Industrial Engineering Dept., METU
Assist. Prof. Dr. Murat Perit Çakır Co-Supervisor, Cognitive Science Dept., METU
Examining Committee Members:
Assoc. Dr. Canan Sepil Industrial Engineering Dept., METU
Prof. Dr. Canan Çilingir
Industrial Engineering Dept., METU
Assist. Prof. Dr. Murat Perit Çakır
Cognitive Science Dept., METU
Assoc. Dr. Sedef Meral Industrial Engineering Dept., METU
Assoc. Dr. Ferda Can Çetinkaya Industrial Engineering Dept., Çankaya University
Date: 24/06/2015
I hereby declare that all information in this document has been
obtained and presented in accordance with academic rules and ethical
conduct. I also declare that, as required by these rules and conduct, I
have fully cited and referenced all material and results that are not
original to this work.
Name, Last Name: Betül Batun
Signature:
v
ABSTRACT
GENDER AND PERSONALITY DIFFERENCES IN COGNITIVE TASKS:
A PERFORMANCE AND HUMAN MENTAL WORKLOAD STUDY
Batun, Betül
M.S., Department of Industrial Engineer
Supervisor: Prof. Dr. Canan Çilingir
Co-Supervisor: Assist. Prof. Dr. Murat Perit Çakır
June 2015, 154 Pages
The increasing complexity of today’s systems creates more demand on
individuals during their daily tasks and mental workload becomes more
important during a cognitive process. In this study, our aim is to investigate
whether individual differences, cognitive task types and task difficulty have
a significant impact on mental workload and how those elements affect
mental workload if they do. N-back, maze and information sampling task
(IST) are used as cognitive tasks. Mental workload is assessed by monitoring
changes in blood oxygenation using functional near-infrared spectroscopy
(fNIR). It is observed that males do more trials than females during IST,
males have better 4-back performance but no gender differences are
observed for other performance measures. No significant difference between
genders exists in blood oxygenation levels. Agreeableness and
conscientiousness personality traits have negative correlations with IST
performance. Extraversion, agreeableness, conscientiousness and openness
to experience have correlations with blood oxygenation levels. Lastly, we
observe from our results that performance decreases with the increasing
vi
difficulty level and oxygenation level in voxels increases with the increasing
difficulty level. In the future, a larger sample of participants, more difficult
tasks and participants from different fields and schools should be chosen in
order to provide more reliable results.
Keywords: Cognitive task, mental workload, gender differences, personality
traits, fNIR
vii
ÖZ
BİLİŞSEL GÖREVLERDE CİNSİYET VE KİŞİLİK ÖZELLİKLERİ
FARKLILIKLARI: PERFORMANS VE ZİHİNSEL İŞ YÜKÜ
ÇALIŞMASI
Batun, Betül
Yüksek Lisans, Endüstri Mühendisliği
Tez Yöneticisi: Prof. Dr. Canan Çilingir
Ortak Tez Yöneticisi: Yrd. Doç. Dr. Murat Perit Çakır
Haziran 2015, 154 Sayfa
Günümüz sistemlerinin karmaşıklığı günlük görevleri sırasında bireylerden
daha fazla yararlanmaya yönelik bir sistem yaratmış ve bu da karşımıza
dinamik bir yapıya sahip olan zihinsel iş yükünü çıkarmıştır. Bu çalışmadaki
hedefimiz, cinsiyetin ve kişilik faktörlerinin, bilişsel görev tipinin, görev
zorluk derecesinin performansa ve zihinsel iş yüküne olan etkisini
araştırmaktır. N-geri, labirent ve bilgi seçme görevleri bu çalışmada bilişsel
görevler olarak kullanılmıştır. Zihinsel iş yükü ölçümü için fonksiyonel yakın
kızılötesi spektroskopi (fNIR) cihazı kullanılmıştır. Sonuçlara göre erkekler
bilgi seçme görevinde kadınlardan daha çok deneme yapmışlar ve erkekler 4-
geri görevinde kadınlardan daha iyi bir performans göstemişlerdir. Diğer
görevlerdeki performanslarda ise cinsiyet farkına rastlanmamıştır. Ayrıca
kanlanma değerleri arasında farklılık olmaması, bilişsel görevlerde kadın ve
erkeğin zihinsel yorgunlukları arasında bir fark olmadığını göstermektedir.
Kişilik faktörlerinin performansla ilişkisine bakıldığında, uyumluluk ve
sorumluluk faktörleri ile bilgi seçme performans ölçümleri arasında negatif bir
viii
korelasyon gözlemlenmiştir. Dışadönüklük, uyumluluk, sorumluluk ve
deneyime açık olma faktörlerinin kanlanma değerleriyle korelasyonları vardır.
Bir başka amaç da performansı ve kanlanma seviyelerini farklı zorluk
derecelerinde incelemektir. Sonuçlara göre, beklenildiği gibi, zorluk seviyesi
arttıkça kişilerin doğru yapma seviyeleri azalmakta, kanlanma seviyeleri
artmaktadır. İleriki çalışmalarda daha geniş bir örneklem, daha zor bilişsel
görevler, daha farklı okullardan ve daha farklı bölümlerden katılımcıların
kullanılması daha güvenilir sonuçlar sağlayacaktır.
Anahtar kelimeler: Bilişsel görevler, zihinsel iş yükü, cinsiyet farkı, kişilik
farkı, Fnir
ix
To my grandmother
x
ACKNOWLEDGEMENTS
First and foremost, I would like to express the deepest appreciation to my supervisor
Prof. Dr. Canan Çilingir for the valuable guidance, moral support and friendly
tolerance throughout this study. She inspired me greatly to work in this study and
she was much more than an advisor to me. I cannot thank her enough for this great
experience.
I would like to give special thanks to Assist. Prof. Dr. Murat Perit Çakır, my co-
supervisor for his valuable supervision and insightful comments in every step of the
study. I am grateful to him for his guidance and positive attitude.
I would also thank to Assoc. Dr. Canan Sepil, Assoc. Dr. Sedef Meral and Assoc.
Dr. Ferda Can Çetinkaya for them being in my examining committee and their
valuable contributions to this study.
I cannot thank my sister Sakine Batun enough. Without her support, understanding,
love and friendship this work could not be completed. Her contribution means more
than I could express just by words.
I am beyond grateful to my parents, Ahmet Batun and Ayşe Batun who supported
me along the way, making this effort possible. For the last three years, I have been
devoted to this thesis study, which has been hard work but finally paid off. I am
grateful to my family, my friends and all those who accompanied and supported me
during this journey.
Last but not least, I would like to thank all participants in the experiments for their
time and contribution.
xi
TABLE OF CONTENTS
ABSTRACT ..................................................................................................................... v
ÖZ ................................................................................................................................... vii
ACKNOWLEDGEMENTS ............................................................................................. x
TABLE OF CONTENTS ................................................................................................ xi
LIST OF FIGURES ....................................................................................................... xiv
LIST OF TABLES ........................................................................................................ xvi
CHAPTERS ...................................................................................................................... 1
1. INTRODUCTION ........................................................................................................ 1
1.1. The Purpose of the Study ............................................................................... 2
1.2. Research Questions ........................................................................................ 3
1.3. Organization of the Thesis .............................................................................. 4
2. LITERATURE REVIEW ............................................................................................. 5
2.1. Mental Workload ............................................................................................ 5
2.1.1. Subjective Ratings .................................................................................. 6
2.1.2. Performance Measures ........................................................................... 6
2.1.3. Pysilogical Measures .............................................................................. 7
2.2. Cognitive Tasks ............................................................................................ 11
2.3. Individual Differences .................................................................................. 14
2.3.1. Gender Differences ............................................................................... 14
2.3.2. Personality Differences ........................................................................ 20
3. THE METHODOLOGY ............................................................................................ 27
3.1. Participants ................................................................................................... 27
xii
3.2. Apparatus ...................................................................................................... 27
3.2.1. Functional Near-Infrared (fNIR) Spectroscopy .................................... 27
3.2.2. Information Sampling Task .................................................................. 29
3.2.3. N-Back Task ......................................................................................... 31
3.2.4. Maze Task ............................................................................................. 33
3.2.4.1. Maze Maker .................................................................................. 34
3.2.4.2. Maze Walker ................................................................................. 35
3.2.4.3.Maze Analyzer ............................................................................... 37
3.2.5. Basic Personality Traits Inventory ........................................................ 38
3.3. Experimental Procedure ............................................................................... 38
4. RESULTS AND DISCUSSIONS ............................................................................... 41
4.1. Data Analysis ................................................................................................ 41
4.2. Gender Difference in Performance Measures ............................................... 44
4.3. Task Level Differences in Performance Measures ....................................... 55
4.4. Gender Differences in fNIR Measures ......................................................... 63
4.5. Task Level Differences in fNIR Measures ................................................... 69
4.6. The Relationship of fNIR Measures ............................................................. 84
4.7. Correlation between fNIR and Performance Measures ................................ 87
4.8. Personality Traits .......................................................................................... 93
4.9. Personality Differences in Performance Measures ....................................... 95
4.10. Personality Differences in fNIR Measures ............................................... 106
5. CONCLUSION ......................................................................................................... 117
5.1. Outcomes of the Study ................................................................................ 117
5.2. Limitations of the Study.............................................................................. 120
REFERENCES ............................................................................................................. 122
xiii
APPENDICES .............................................................................................................. 137
A. APPROVAL BY THE HUMAN SUBJECT ETHICS COMMITTEE .............. 137
B. PARTICIPANTS’ PERSONAL DATA ............................................................. 138
C. PERSONALITY QUESTIONNAIRE ................................................................ 139
D. ENGLISH VERSION OF THE PERSONALITY QUESTIONNAIRE ............ 140
E. THE EVALUATION OF BASIC PERSONALITY TRAITS INVENTORY ... 141
F. INFORMED CONSENT FORM ........................................................................ 142
G. MAZE COMPLETION TIME ........................................................................... 143
H. MAZE PATH LENGTH .................................................................................... 144
I. IST PERFORMANCE ......................................................................................... 145
J. IST COMPLETION TIME .................................................................................. 146
K. IST TRIAL NUMBERS ..................................................................................... 147
L. N-BACK PERFORMANCES ............................................................................. 148
M. FNIR MEASURES DURING MAZE TASK .................................................... 149
N. FNIR MEASURES DURING IST ..................................................................... 150
O. FNIR MEASURES DURING N-BACK TASK ................................................ 151
P. THE PERSONALITY TRAIT STATISTICS ..................................................... 152
R. THE PERSONALITY TRAIT RESULTS ......................................................... 153
S. THE PERSONALITY TRAIT HIGHER/LOWER RESULTS .......................... 154
xiv
FIGURES
Figure 1: fNIR System Design ........................................................................................ 28
Figure 2 - fNIR Sensors Placed on the Participant’s Forehead ...................................... 29
Figure 3: Screen Display for the IST (Clark et al., 2006) ............................................... 31
Figure 4: Screen Display for N-back Task ..................................................................... 32
Figure 5: Design of the Maze in Maze Walker ............................................................... 34
Figure 6: Initial Setting Screen of Maze Walker ............................................................ 36
Figure 7: Maze Analyzer ................................................................................................ 37
Figure 8: Flowchart of Data Analysis ............................................................................. 44
Figure 9: Boxplot of Maze Completion Time according to Gender ............................... 45
Figure 10: Boxplot of Maze Path according to Gender .................................................. 46
Figure 11: Boxplot of IST Completion Times according to Gender .............................. 49
Figure 12: Boxplot of IST Performances according to Gender ...................................... 50
Figure 13: Boxplot of IST Trial Numbers according to Gender ..................................... 52
Figure 14: Boxplot of N-back Performances according to Gender ................................ 55
Figure 15: Boxplot of IST Performances according to Task Difficulty ......................... 56
Figure 16: Boxplot of IST Completion Times according to Task Difficulty ................. 58
Figure 17: Boxplot of IST Trial Numbers according to Task Difficulty ........................ 59
Figure 18: N-back Performance according to Task Difficulty ....................................... 60
Figure 19: Boxplot of Oxygenation Levels at Left DLPFC ........................................... 71
Figure 20: Boxplot of Oxygenation Levels at Right DLPFC ......................................... 72
Figure 21: Boxplot of Oxygenation Levels for Average of Left and Right DLPFC ...... 72
Figure 22: Location of Voxels at Prefrontal Cortex ....................................................... 85
Figure 23: Boxplot of Maze Completion Time ............................................................. 143
Figure 24: Boxplot of Maze Path Length ..................................................................... 144
Figure 25: Boxplot of IST Performance ....................................................................... 145
Figure 26: Boxplot of IST Completion Time ............................................................... 146
Figure 27: Boxplot of IST Trial Numbers .................................................................... 147
xv
Figure 28: Boxplot of N-back Performances ................................................................ 148
Figure 29: Boxplot of Maze fNIR Measures ................................................................ 149
Figure 30: Boxplot of IST fNIR Measures ................................................................... 150
Figure 31: Bocplot of N-back fNIR Measures ............................................................. 151
xvi
TABLES
Table 1: Summary of Studies on the Gender Differences .............................................. 18
Table 2: Summary of Studies on Personality Differences .............................................. 24
Table 3: Normality Tests for fNIR Variables ................................................................. 43
Table 4: Normality Tests for Performance Variables ..................................................... 44
Table 5: Mann Whitney U-Test Summary for Maze Performance Measures ................ 47
Table 6: Mann-Whitney U-Test Summary for IST Performance Measures ................... 51
Table 7: Independent t-Test Summary for IST Trial Numbers....................................... 53
Table 8: Mann Whitney U-Test Summary for 1-back and 2-back ................................. 54
Table 9: t-Test Summary for 3-back and 4-back ............................................................ 54
Table 10: Wilcoxon Signed Ranks Test Summary for IST Performances ..................... 57
Table 11: Wilcoxon Signed Ranks Test Summary for IST Completion Times ............. 58
Table 12: Paired t-Test Summary for IST-1 and IST-2 Trial Numbers .......................... 59
Table 13: Repeated ANOVA Test Summary for N-back Performances ........................ 61
Table 14: Pairwise Comparisons for N-back Performances ........................................... 62
Table 15: Descriptive Statistics for Oxygenation Levels during Maze Tasks ................ 64
Table 16: Independent Samples Test for Maze fNIR Measures ..................................... 64
Table 17: Descriptive Statistics for Oxygenation Levels during ISTs ........................... 66
Table 18: Repeated ANOVA for IST fNIR Measures of Left DLPFC .......................... 66
Table 19: Repeated ANOVA for IST fNIR Measures of Right DLPFC ........................ 66
Table 20: Descriptive Statistics for Oxygenation Levels during N-back Tasks ............. 68
Table 21: Repeated ANOVA for N-back fNIR Measures of Left DLPFC ..................... 68
Table 22: Repeated ANOVA for N-back fNIR Measures of Right DLPFC .................. 69
Table 23: Descriptive Statistics of Oxygenation Levels of Left DLPFC ....................... 70
Table 24: Descriptive Statistics of Oxygenation Levels of Right DLPFC ..................... 70
Table 25: Descriptive Statistics of Oxygenation Levels for Average of Left and
Right DLPFC .................................................................................................................. 71
Table 26: Paired Samples Test to Compare Maze Task with others for Left DLPFC.... 74
xvii
Table 27: Paired Samples Test to compare Maze Task with others for Right DLPFC .. 74
Table 28: Paired Samples Test to compare Maze Task with Others for Average of
Left and Right DLPFC ................................................................................................... 74
Table 29: Paired Samples Test to Compare IST with others for Left DLPFC ............... 77
Table 30: Paired Samples Test to Compare IST with others for Right DLPFC ............. 78
Table 31: Paired Samples Test to Compare IST with others for Average of Left and
Right DLPFC .................................................................................................................. 79
Table 32: Paired Samples Test for compare N-back Tasks with Others for Left
DLPFC ............................................................................................................................ 81
Table 33: Paired Samples Test for compare N-back Tasks with Others for Right
DLPFC ............................................................................................................................ 82
Table 34: Paired Samples Test for compare N-back Tasks with Others for Average
of Left and Right DLPFC ............................................................................................... 83
Table 35: Descriptive Statistics for Oxygenation Levels at Left and Right DLPFC ..... 86
Table 36: Paired t-Test for Oxygenation Levels at Left and Right DLPFC ................... 86
Table 37: Correlations between Maze Measures ............................................................ 87
Table 38: Correlations between IST-1 Measures ........................................................... 88
Table 39: Correlations between IST-2 Measures ........................................................... 89
Table 40: Correlations between 1-back Measures .......................................................... 90
Table 41: Correlations between 2-back Measures .......................................................... 91
Table 42: Correlations between 3-back Measures .......................................................... 92
Table 43: Correlations between 4-back Measures .......................................................... 93
Table 44: Descriptive Statistics for Personality Traits in Turkey (Gencoz and Oncul,
2012) ............................................................................................................................... 95
Table 45: Higher and Lower Classification for Personality in the Experiment ............. 95
Table 46: Test Summary of All Personality Traits for Maze Performance Measures ... 97
Table 47: Group Statistics of All Personality Traits for Maze Performance Measures . 98
Table 48: Test Summary of All Personality Traits for IST Performance Measures .... 100
Table 49: Group Statistics of All Personality Traits for IST Performance Measures .. 101
Table 50: Test Summary of All Personality Traits for N-back Performance Measures103
Table 51: Group Statistics of All Personality Traits for N-back Performance
xviii
Measures ....................................................................................................................... 104
Table 52: Test Summary of All Personality Traits for fNIR Measures of Left
DLPFC .......................................................................................................................... 107
Table 53: Test Summary of All Personality Traits for fNIR Measures of Right
DLPFC .......................................................................................................................... 108
Table 54: Group Statistics of Extraversion Personality Trait for fNIR Measures ........ 109
Table 55: Group Statistics of Conscientiousness Personality Trait for fNIR Measures110
Table 56: Group Statistics of Agreeableness Personality Trait for fNIR Measures ..... 111
Table 57: Group Statistics of Neuroticism Personality Trait for fNIR Measures ......... 112
Table 58: Group Statistics of Openness to Experience Personality Trait for fNIR
Measures ....................................................................................................................... 113
Table 59: Group Statistics of Negative Valence Personality Trait for fNIR Measures 114
Table 60: Descriptive Statistics of Personality Traits ................................................... 152
Table 61: Detailed Results of the Personality Trait ...................................................... 153
Table 62: Higher/Lower Results of the Personality Trait ............................................. 154
1
CHAPTER 1
INTRODUCTION
Recent advances and the increasing complexity of today’s technology create more
demand on individuals while performing everyday tasks. This increasing demand
becomes more difficult to be satisfied and brings an excessive burden over limited
information processing capacity of a person. Especially in complex systems based
on human control in time-sensitive contexts, the human factor should be considered
as an important factor. For example, air traffic control (ATC) to maintain traffic
safety and efficiency, and human operations in health care are examples of high
hazard industries that need to be considered from a cognitive processing capacity
perspective.
Mental workload is the operator’s excess capacity between demands of the task and
his/her current capacity limits while achieving adequate task performance (Jex,
1988). Human mental workload is encountered in many complex systems. For
example: Air traffic controllers experience mental workload while remembering
positions and heights of airplanes from the ground level. Any complex or long
surgery forces a doctor’s mental strain. Long hours of meeting in management based
systems can generate high levels of cognitive workload. Examples of human mental
workload are easy to observe in any complex system.
2
Human mental workload, which is the margin between the current task demand and
the operator’s motivated capacity, has a dynamic structure (Jex, 1988). The
complexity of a task may change according to the task type, the capability of
performing a task may differ depending on the variety of people and both have
impacts on human mental workload. Individual differences, cognitive task types and
task difficulty are thought as they have an effect on mental workload. However,
whether these factors have really an effect on mental workload and how these
elements affect mental workload if they do, are open questions waiting to be
answered. This is the main motivation behind our study.
Functional near infrared spectroscopy (fNIR) is used within the study as a
physiological method to assess mental workload. Studies show that human brain
hemodynamic activity assessed by fNIR can provide a sensitive and reliable
assessment of human mental workload during a complex task (Chance et al., 1998;
Izzetoglu et al. 2004; Ayaz et al., 2008). In this study, we used fNIR as a measure
tool of mental workload and tried to explain relationships between mental workload
and individual differences of users.
1.1. The Purpose of the Study
The aim of this study is to investigate the effects of gender on different cognitive
tasks by measuring performance and monitoring changes in blood oxygenation using
fNIR to assess mental workload. Another purpose is to find out if there are any
personality differences on performances and blood oxygenation levels during
cognitive tasks. The findings for these aims will help literature about how individual
differences affect mental workload by investigating both gender and personality
differences.
This study investigates whether performance and fNIR measures change according
to the task difficulty level. Increase in the difficulty of the task may cause reduction
in performance and rising activation in frontal cortical areas. Difficulty degrees of
3
different cognitive tasks are explored by fNIR measures. Since fNIR has been used
to assess mental workload, various types of cognitive tasks may cause different
levels of fatigue according to their difficulty. The results will provide information
about comparison of different kinds of cognitive tasks. In the literature, comparing
both levels of cognitive tasks and different task types does not exist. Therefore, this
study will contribute.
1.2. Research Questions
Research questions of this study are determined according to the purpose of the
study. The following research questions are investigated through experiments
conducted in this study:
Do performance levels of subjects change according to gender, when the same
cognitive tasks are performed?
Do performance levels of subjects change according to gender at different levels
of the same cognitive tasks?
Do performance levels of subjects change according to the difficulty levels of
the same cognitive tasks?
Do changes in blood oxygenation levels of subjects vary according to gender,
when the same cognitive tasks are performed?
Do changes in blood oxygenation levels of subjects vary according to gender at
different levels of the same cognitive tasks?
Do changes in blood oxygenation levels of subjects change according to the
difficulty levels of the same cognitive tasks?
Is there a difference between changes in oxygenation levels of left dorso-lateral
prefrontal cortex (DLPFC) and right DLPFC, during a cognitive task?
Do changes in blood oxygenation levels of subjects change according to type of
the cognitive tasks?
4
Is there a relationship between changes in blood oxygenation and performance
level during a cognitive task?
Do performance levels of subjects change according to personality traits, when
the same cognitive tasks are performed?
Do performance levels of subjects change according to personality traits at
different levels of the same cognitive tasks?
Do changes in blood oxygenation of subjects change according to personality
traits, when the same cognitive tasks are performed?
Do changes in blood oxygenation of subjects change according to personality
traits at different levels of the same cognitive tasks?
1.3. Organization of the Thesis
This chapter explains the aim of the study and presents the research questions.
Chapter 2 presents a summary of the literature related to this study. Chapter 3 gives
information about the experimental procedure including the participants, the
apparatus, the experimental process and the cognitive tasks used. Chapter 4 presents
the results of the statistical analysis and provides a general discussion of the
findings. Chapter 5 includes conclusions, limitations and recommendations.
5
CHAPTER 2
LITERATURE REVIEW
In this chapter, mental workload and its measure techniques will be briefly
reviewed. Then, types of cognitive tasks, gender differences and lastly personality
differences during cognitive tasks will be explained.
2.1. Mental Workload
A generally accepted definition of mental workload does not exist, because it covers
a broad range of factors. Jex (1988, p.9) defines mental workload as “… the
operator’s evaluation of the attentional load margin (between their motivated
capacity and the current task demands) while achieving adequate task performance
in a mission-relevant context.” De Waard (1996) offers an alternative definition of
mental workload by saying that it is the specification of the amount of information
processing capacity that is used for task performance. Gopher & Braune (1984)
stated that mental workload can be defined as a mental strain resulting from
performing a task. It is a concept to explain the inability of human operators to cope
with the requirements of a task. Mental workload can vary by task complexity,
motivation and fatigue level of the operator. Therefore, it can be said that mental
workload is both task- and person-specific. These factors change mental workload,
but do not change its definition.
6
Measuring mental workload may provide insights about the design of the system,
the possible sources of human errors, and the reasons underlying environmental
problems such as noise factors, luminance. Such insights can be used to design
better systems. Three main categories of mental workload measure are defined in the
literature: subjective measures, performance measures, and physiological measures
(Eggemeier, 1988; De Waard, 1996). Each measure is defined and explained under
its own title in the following paragraphs.
2.1.1. Subjective Ratings
In this technique, the participant either answers a questionnaire or rates his/her
perceived mental workload in a specific situation. If subjective ratings are well-
structured, they may be helpful in learning the situation of the participant directly.
There are many subjective assessment methods. The frequently used ones are as
follows: Bedford scales (Wierwille and Casali, 1983), Overall Workload (Vidulich
and Tsang, 1987), NASA TLX (Hart and Staveland, 1988), SWAT (Reid and
Nygren, 1988), and W/Index (Workload Index: North and Riley, 1989). Although
they are cheap and easy to use, the repeatability and validity of these instruments
can be uncertain. To get around this problem, other methods are often used together
with subjective ratings.
2.1.2. Performance Measures
Performance measures of workload can be grouped into two types: primary task
measures and secondary task measures. Primary task measure is based on measuring
the performance observed during the primary task, but the performance level alone
is not considered as an indicator of mental effort because of strategic reallocation of
mental capacity (Wilson, 2004). Thus, simply considering performance to assess the
participant’s mental effort may not be sufficient. There are many primary task
measures and the most frequently used ones are as follows: Accuracy, number of
errors, reaction times, and speed (O’Donnell & Eggemeier, 1986).
7
Secondary task measure is based on measuring the performance of a secondary task
while the participant is performing a primary task. For example, performing basic
arithmetic calculations during driving and talking to other pilots during flight are
secondary tasks for a driver and a pilot, respectively. The secondary task measure is
classified into two methodologies: Auxiliary Task and Loading Task (O’Donnell
and Eggemeier, 1994). Secondary task measure is less frequently practiced as
compared to subjective ratings. Colle and Reid (1999) claimed that globally
sensitive secondary task measure methods are needed. However, different primary
tasks require different secondary tasks, which create generalizability issues
(Meshkati & Lowewinthal, 1988).
2.1.3. Physiological Measures
Physiological measure is an objective measure approach based on physical reactions
of the participant’s body as monitored by specific sensors. These physical reactions
can be monitored in terms of skin conductivity, cardiovascular activity, respiratory
responses, brain activity or pupillary size, each of which are measured by a
dedicated equipment specialized on that modality (Wilson & O’Donnell, 1988).
Heart rate (HR) and heart rate variability (HRV) are the most commonly used
measures of workload, because the heart beat is a relatively easy measure to obtain
(Wilson and O’Donnell’ 1988). Generally, increased HR is associated with
increased workload (Jorna, 1993; Roscoe, 1993; Wilson, Fullenkamp & Davis,
1994; Hankins & Wilson, 1998), but some studies argued that there are other
psychological, physical and environmental factors that could affect HR (Lee &
Parks, 1990; Roscoe, 1992). Therefore, to measure mental workload with
increased/decreased HR could be misleading.
Some studies showed that heart rate variability (HRV) decreases under high
workload levels (Mulder, 1979; Vincente, Thornton & Moray, 1987; Mulder &
Mulder, 1997), but there are also studies that reported conflicting results where
8
increased workload did not lead to a decrease in HRV (Brookings, Wilson, & Swain,
1996; Hankins & Wilson, 1998). Therefore HRV is not a totally reliable measure for
cognitive workload. Like HR, HRV may be affected by not only mental workload
but also external factors.
The electroencephalogram (EEG) is a workload measure that uses surface electrodes
placed on a participant’s scalp to monitor the changes in electric potentials due to
the brain’s activity during a specific task. (DE Waard, 1996). EEG focuses on brain
activity of a person. There are some types of bands (delta, theta, alpha, beta and ultra
beta) that are used with this method to measure changes in the electrical potential in
the brain cells (Brookings et al., 1996). Simulation based studies (Borghini et al.,
2011; Dussault et al., 2005) showed that there is a power increase in theta band and
changes in alpha band with low performance. Electrons on front-middle and top-
middle also showed a power increase in theta band with distraction due to task
difficulty. However, physical movements of the subjects can be a problem with
this method. Physical limitations, non-practical structure and the cost of
implementation of EEG can cause a problem during application.
Electrooculogram (EOG) measures the velocity of the saccadic eye movements to
find workload (Galley, 1993). Borgini et al. (2014) used EOG in their study. Their
results showed when there was an increase in the workload of the pilot; increases in
frequency and length of blinking, a reduction in focusing were observed. However,
there are external factors such as the amount of light in the cockpit that can also
change the pupil size and blink frequencies (De Rivecourt et al., 2008). In short,
EOG offers good indicator for visual workload, but it is not enough to see total
cognitive workload.
Positron Emission Tomography (PET) is an important imaging technique for
measuring blood flow or energy consumption in the brain. PET is an invasive
technique that requires the injection of harmless radioactive tracers in the blood
9
stream. The half-life of the tracer limits the PET experiments to relatively small
durations, and the subjects are required to stay in a confined position to ensure
accurate detection of brain activity. Single Photon Emission Computed
Tomography (SPECT) is another brain imaging technique which is similar to PET,
but it involves less detail and is less expensive than PET (Greely & Wagner, 2011).
The image quality in SPECT makes functional brain analysis challenging.
Functional magnetic resonance imaging (fMRI) is another popular and common
neuroimaging technique. It allows detecting increases in blood oxygen levels and
shows detailed maps of the brain areas underlying human mental activities (Greely
& Wagner, 2011). It is very useful to detect complex emotions in the brain, but it is
costly and hard-to-use as a neuroimaging tool (Cansiz, 2012). Moreover, it requires
subjects to be monitored in an MRI scanner in a confined position, which makes it
difficult to conduct practical studies in human factors research.
Functional near-infrared spectroscopy (fNIR) is a relatively new neuroimaging
technology compared to commonly employed techniques such as EEG, PET,
SPECT and fMRI. It has been used in functional brain studies to monitor changes in
the concentration of oxygenated hemoglobin (HbO) and deoxygenated hemoglobin
(HbR).
After Chance’s (1998) pioneering work on the development of fNIR technology and
his demonstration of its use for monitoring oxygenation changes in the brain due to
the modulation of specific cognitive processes, many follow-up studies consistently
produced compatible results for the monitoring of mental workload with fNIR.
Izzetoglu et al. (2004) used fNIR to help to understand the cognitive state of a user
during a complex task. The complex task was a video game called the Warship
Commander Task (WCT). They used different task difficulty and task load levels by
changing the state of the game. The main aim of the study was to see that fNIR was
an appropriate measure to predict changes in cognitive workload. Therefore, the
10
hypothesis of the study was that blood oxygenation in the prefrontal cortex which
was assessed by fNIR, would have a raise with increasing task difficulty and have a
positive correlation with performance. When a task became too difficult, the
participant was expected to disengage his/her attention on the cognitive task and the
blood oxygenation level was expected to decrease. Results showed that change in
the blood oxygenation was significantly sensitive to task difficulty and had a
positive correlation with performance. So, it indicated that fNIR could be reliably
used to monitor hemodynamic changes in the dorsolateral prefrontal cortex in a
complex working memory and decision-making task.
Another study whose aim was to measure changes in mental workload at different
difficulty levels by using fNIR was done by Ayaz et al. (2012). They used the N-
back short term memory test and an air traffic controller simulation task in their
study. The authors observed a monotonic decrease in reaction time and accuracy as
the degree of n in n-back task was increased. fNIR results were also sensitive to this
change, especially at left interior frontal gyrus. The mean oxygenation level in the
arterior medial frontopolar cortex had a monotonic increase with increased task
difficulty in the complex task (air traffic control task). Brain activities on different
areas of prefrontal cortex according to different types of tasks (working memory and
planning/decision making) were observed by fNIR. Ayaz et al. (2012) claim that
this finding parallels with fMRI findings on the differentiation of lateral and medial
prefrontal cortex in comparable tasks. To sum up, fNIR is an optical brain imaging
technology that can be reliably used to measure hemodynamic changes in the
prefrontal cortex and its relations with mental workload, expertise and performance
in different tasks. fNIR results are in the agreement with subjective measures and
earlier neuroimaging studies (Chance et al., 1998; Bunce et al., 2006; Coyle et al.,
2007 and Izzetoglu et al., 2004).
When we compare fNIR with other neuroimaging techniques, we find both
advantages and disadvantages of them. Firstly, EEG is relatively inexpensive,
11
noninvasive; its results need reconstruction and can only provide a relative
approximation. fMRI is safe and noninvasive neuroimaging with high spatial
resolution. fMRI and fNIR are both indirect measures of neural activity limited by
hemodynamic response. They are both safe and provide a level spatial resolution,
but fNIR cannot substitute fMRI. When fMRI has a better spatial resolution, fNIR is
limited to the outer cortex while other techniques can provide a richer picture of
brain (Bunce et al., 2006). On the other side, fNIR tolerates movements for
participants; it is more comfortable than other technologies. fNIR is safe, highly
portable, user-friendly and relatively inexpensive, with rapid application times and
near-zero run-time costs and these features make it a suitable tool for brain imaging
in many operations compared to other brain activity measure (Izzetoglu, 2004).
fNIR can be helpful to understand the cognitive and emotional state of the user
during mentally demanding operations.
2.2. Cognitive Tasks
Cognitive ability is an executive function to organize a sequence of actions toward a
goal (Fuster, 2008). Cognitive abilities that a brain performs are various and hard to
categorize. Many studies show many different classifications. According to Deco
and Rolls (2005); memory, attention and decision-making are three fundamental
functions. Fuster (2008) had a longer and detailed list as following: attention,
memory, working memory, planning, temporal integration, decision-making,
monitoring, and inhibitory control. In our study, we followed up Deco and Rolls
(2005) categorization.
Attention is a selection process by maintaining a certain amount of information from
environment (Smith & Kosslyn, 2009). Working memory is a limited capacity
system that involves storage, processing and manipulation of information and has an
ability to encode new information into long-term memory storage (Johnson, 1992).
Memory and attention are related to each other. In the industry, most employees use
12
these functions during their daily working process. For example, air traffic
controllers need to remember positions and altitudes of airplanes to prevent possible
aircraft accidents. Pilots need to pay attention to every signal on display while
flying. Drivers have to remember traffic signs on their road, pay attention to
pedestrians and control their signals. Cognitive tasks that include memory can come
up in the healthcare sector. Doctors and nurses have to remember blood pressure and
other vital charts of a patient and take actions according to these vital numbers
during surgery.
The Paced Auditory Serial Addition Task (PASAT; Gronwall, 1977) is a task to
measure working memory and attention capacity. During the task, numbers are
presented one by one every three seconds and the participants are asked at each step
to sum up the last two numbers shown. This task involves both a mathematical
procedure (summing) and remembering the numbers.
Operation Span, Reading Span and Counting Span, are complex span measures
which are widely used for measuring working memory capacity (Daneman &
Carpenter, 1980; Unsworth et al., 2005). In the Operation Span task, people have to
both solve algebraic operations and remember unrelated words (Turner & Engle,
1989). In the Reading Span Task, sentences are presented one by one and subjects
have to remember last word of each sentence and at the end of the task, they have to
say all the last words in the presented order (Daneman & Carpenter, 1980). In the
Counting Span task, people have to remember dots presented and count totals (Case
et al., 1982).
N-back task is more commonly used for working memory studies (Carlson et al.,
1998; Gevins et al., 1996). Participants have to decide whether a currently presented
visual stimulus matches the stimulus previously presented “n” trials back during the
task. N could be equal to 0, 1, 2, 3, 4, and more. As n increases, the task becomes
more difficult (Parmenter et al., 2006). Several working memory studies, especially
13
neuroimaging researches, use n-back task to see the role of ventrolateral and
dorsolateral prefrontal cortex (Braver et al., 2001; Manoach et al., 1997; Ragland et
al., 2002).
Decision making is a cognitive process which involves choosing one option from
among a set of alternatives based on a person’s criteria (Wang & Ruhe, 2007).
Simon (1977) divided decision making process into three steps: intelligence
(identification of all possible alternatives), design (determining results of these
alternatives) and choice (evaluate these results). In the first step, a person identifies
the problem, gathers information about the problem and analyses the situation. Then
she/he generates possible alternatives. In the second step of decision making
process, the person evaluates the results of possible options. In the third and last step
she/he evaluates alternatives and selects a best alternative for her/him. In industry, it
generally shows up at management departments of every sector or managerial-based
sectors. For example, a director has to make decisions every day by considering his
company’s benefits. Doctors have to decide which medicine is better for a patient,
planning engineers have to decide how many products must be produced next
month, and logistic engineers have to decide where to establish their warehouses.
Some examples of the cognitive tasks used in the studies involving brain imaging
techniques are as follows: Mandrick et al. (2013) used fNIR to see how an additional
mental load affects brain activation and used an arithmetic task as a cognitive task.
Cui et al. (2011) compared fNIR to fMRI for monitoring brain function across
multiple cognitive tasks. They used four types of tasks: finger tapping, go/no-go
task, judgment of line orientation task and n-back task. Bell et al. (2005) studied
gender differences in brain activities during cognitive tasks by using fMRI. They
considered a variety of cognitive task types to ensure reliable information on gender
differences. These were verbal fluency, spatial attention, working memory and
motor tasks. If we consider studies that used fNIR, their cognitive tasks were as
follows: Bunce (2006) used target categorization, Leon-Carrion et al. (2010)
14
selected Luria’s Memory Word Task, Izzetoglu (2011) assigned target
categorization as an attention task and n-back as a working memory task.
2.3. Individual Differences
2.3.1. Gender Differences
There are many studies that use brain imaging technique to investigate gender
differences. Fujimoto et al. (2008) investigated relative metabolic changes due to
age- and gender-related differences in the brain by using PET and MRI analysis.
Their brain research was comprehensive; it included frontal, primary sensorimotor,
parietal, occipital and temporal lobes. Since we discuss frontal cortex activities in
this study, we focused on the results related to these areas. Relative metabolic values
in the frontal lobe showed significant age-related differences between subjects’ at
20’s and 40’s (Their metabolic value decreases by age), but there is no significant
age-related difference between subjects’ at 50’s and 70’s. Significant gender
differences were not apparent in the frontal region in each age interval. Willis et al.
(2002) aimed to determine the effects of age, sex and laterality on cerebral glucose
metabolism (CMRglc) by using PET. The results showed that there was an inverse
relationship between absolute regional CMRglc and age across widespread cortical
areas including frontal cortex. Men had lower absolute metabolism than women
bilaterally in the medial frontal gyrus. Both men and women showed left greater
than right regional CMRglc in the medial frontal cortex while they showed right
greater than left regional CMRglc in lateral frontal cortex. Marumo (2009) used an
emotional activation task to see gender difference in prefrontal cortex by using
fNIR. Females had significantly increased oxy-Hb change relative to males in the
right ventrolateral prefrontal cortex during the latter half of the task period and this
showed that the gender had an effect on individual variability of NIRS signals in
response to emotional stimuli. Yang (2007) searched for gender differences in
prefrontal area activation during emotional stress by using fNIR. The oxy-Hb
15
response in the prefrontal cortex induced by emotional stress of females was
significantly higher than males.
Gender differences in spatial skills is a popular topic which has been investigated by
many studies. Results of these studies vary (Coluccia & Louse, 2004). There are
many previous studies indicating that males show better performance than females
on a diverse set of spatial tests. Galea & Kimura (1993) made an experiment about
tracing a novel route through the town to find out differences between male and
female subjects. Results showed that males made fewer errors and required fewer
trials to learn the novel route than did females. Persson et al. (2013) searched for sex
differences in spatial memory by using three-dimensional virtual mazes and found
that men outperformed women on the maze task. Moffat et al. (1998) investigated
sex differences in spatial ability by using computer-generated mazes. There was a
significant main effect of gender. Moffat et al. (1998) found that males solved the
mazes significantly faster than females across all five trials and males made
significantly fewer errors than females. However, there are also studies that gender
differences in spatial ability are totally absent. O’Laughlin & Brubaker (1998) used
mental rotation and cognitive mapping tasks and found men and women performed
equally well on those tasks. Taylor & Tversky (1992) also used a mental rotation
test and there were no gender differences. Gender differences in visual-spatial
navigation are also hot topics in neuroimaging studies. Gron et al. (2000) did an
experiment with a maze navigation task and found out males had activations in left
hippocampus whereas females had activations in right parietal and right prefrontal
cortex.
There are also studies that employ the n-back task as a working memory cognitive
task to investigate gender differences. Different versions of n-back task are available
such as spatial, verbal and auditory. Li et al. (2010) found no gender differences in
response time and accuracy during verbal 1-back, 2-back and 3-back tasks whereas a
significant gender difference showed up in brain activations. Females had lower
16
amplitudes and were more spatially focused than males. Koch et al. (2007)
compared performances on verbal 0-back and 2-back tasks between males and
females and observed no gender differences in both tasks. Results of Schmidt et al.
(2009) also show that there were no significant gender differences at verbal 0, 1, 2
and 3-back tasks. Voyer et al. (1995) examined cognitive gender differences in the
context of spatial abilities by doing a meta-analysis of many studies and they found
that males outperformed females for spatial-based working memory tasks.
Speck et al. (2000) examined gender differences in brain activation during working
memory tasks by using fMRI. 1-back and 2-back tasks were used as working
memory tasks. Activation of the lateral prefrontal corticex (LPFC), the parietal
corticex (PC) and caudate were observed in both sexes. Specifically, males showed
right-sided dominance while females showed activation in the left hemisphere.
Performance data showed that females had significantly higher accuracy and slightly
slower reaction times. Both groups showed grater left inferior frontal, superior
parietal and middle temporal gyrus activation. Li et al. (2010) worked gender
differences in working memory during verbal n-back tasks by using 16 channel
fNIR. They looked for changes in the concentrations of oxy-hemoglobin, deoxy-
hemoglobin and total hemoglobin. While changes in oxy-hemoglobin and total
hemoglobin exhibit significant gender difference, results showed that changes in
deoxy-hemoglobin did not exhibit a significant gender difference, whereas females
show left-lateralized activation and males showed bilateral activation for changes in
oxy-hemoglobin and total hemoglobin. However, there are also studies (Schmidt,
2009; Haut & Barch 2006) which examined gender differences during n-back task
and found that men and women showed similar brain activity during n-back tasks by
using fMRI.
Gender differences during decision-making involves a wide concept since decision-
making is a complex cognitive process including evaluation, prediction, anticipation
and response. So, many different types of tasks were used as a decision making task
17
in previous studies. Iowa Gambling Task (Bechara et al., 1994), Information
Sampling Task (Clark et al., 2006), Cambridge Gambling Task (Manes et al., 2002)
are the most known decision making tasks. There are also studies that investigate
gender differences during decision making. Several studies found out that males
showed better performance than females during a decision-task (Reavis & Overman,
2001; Bolla et al., 2004). There are also studies which indicated gender differences
during a decision task and their results showed that behavior and earnings were
similar for males and females (Crone et al., 2003; Lighthall et al., 2012). Van der
Bos et al. (2012) also reached the same conclusion that there were no gender
differences on deliberation time, impulsivity or completion time during a decision
task.
fMRI studies on decision-making indicate that ventromedial prefrontal cortex,
dorsolateral prefrontal cortex and lateral frontopolar areas show activations during
decision making processes (Monchi et al. 2001; O’Doughtery et al., 2001). Bolla et
al. (2004) investigated brain activity in the orbitofrontal cortex (OFC) and
dorsolateral prefrontal cortex (DLPFC) during Iowa Gambling Task by using PET
scans and found that men showed greater lateralized brain activity to the right
hemisphere than females, while females showed grater brain activity in the left
DLPFC. This suggests that women may have different cognitive strategies from men
during a decision-making task.
Table 1 summarizes the literature review on gender differences.
18
Table 1: Summary of Studies on the Gender Differences
Study N Females Males Difference Cognitive Task Measurement Analyses Findings
Fujimoto et al.
(2008)
126 62 64 Age, gender Not stated Brain activations
by PET, MRI
t-test, regression
analysis and
Pearson's
correlation
Significant gender
difference, significant
age difference between
20's and 40's, no
significant age
difference between 50's
and 70's
Willis et al.
(2002)
66 28 38 Age, gender Not stated Brain activations
by PET
ANCOVA,
Pearson's
correlation
There was an inverse
relationship between
absolute metabolism and
age and males had lower
absolute metabolism in
the medial frontal gyrus
Marumo (2009) 20 10 10 Gender An emotional
activitation task
Brain activations
by fNIR
t-test, Pearson's
correlation
Females had increased
oxy-Hb change relative
to males in the right
ventrolateral prefrontal
cortex
Yang (2007) 30 19 11 Gender An emotional
stress task
Brain activations
by fNIR
ANOVA with
repeated
measures
Females had increased
oxy-Hb change relative
to males in the prefrontal
cortex
Galea &
Kimura (1993)
97 48 49 Gender Tracing a novel
route
Number of errors,
number of trials
ANOVA with
repeated
measures
Males made fewer
errors and required
fewer trials to learn the
novel route
Persson et al.
(2013)
24 12 12 Gender Maze tasks Pointing errors,
navigation time,
pointing time,
number of mazes
completed, brain
activations by
fMRI
ANOVA with
repeated
measures
Males were more
accurate at pointing,
males were faster to
complete the maze tasks,
males had greater brain
activitations in the right
posterior and anterior
hippocampus
Moffat et al.
(1998)
74 34 40 Gender A maze task,
spatial ability tests
and verbal ability
tests
Performance,
number of errors,
completion time
ANOVA with
repeated
measures
Males solved the maze
faster and made fewer
errors during the maze,
males had better
performance on spatial
ability tests, no sig.
gender difference on
verbal ability tests
O’Laughlin &
Brubaker
(1998)
160 82 78 Gender A mapping task Accuracy 2-factor ANOVA No sig. gender
difference
Taylor &
Tversky (1992)
70 Not
stated
Not
stated
Gender A mental rotation
test
Performance Not stated No sig. gender
difference
Gron et al.
(2000)
24 12 12 Gender Maze tasks Brain activations
by fMRI
t-test Males had activations in
left hippocampus,
females had activations
in right parietal and right
prefrontal cortex
19
Table 1 (continued): Summary of Studies on Gender Differences
Study N Females Males Difference Cognitive Task Measurement Analyses Findings
Li et al. (2010) 54 28 26 Gender Verbal 1-back, 2-
back and 3-back
tasks
Accuracy,
response time and
brain
acitivitations by
fNIR
ANOVA with
repeated
measures
Females show left-
lateralized activation
and males showed
bilateral activation for
changes in oxy-
hemoglobin and total
hemoglobin
Koch et al.
(2007)
40 19 21 Gender Verbal 0-back and
2-back tasks
Performance,
brain activiations
by fMRI
ANOVA with
repeated
measures,
ANCOVA, t-test
No gender difference in
performance, females
showed stronger brain
activations in a
widespread network
than males
Schmidt et al.
(2009)
50 Not
stated
Not
stated
Gender Verbal 0-back, 1-
back, 2-back and 3-
back tasks
Performance,
brain activiations
by fMRI
ANOVA with
multivariate
repeated
measures
No sig. gender
difference
Speck et al.
(2000)
17 8 9 Gender Verbal 1-back and
2-back tasks
Brain activations
by fMRI,
response time and
accuracy
2-way ANOVA Males showed right-
sided dominance while
females showed
activation in the left
hemisphere, females had
significantly higher
accuracy and slightly
slower reaction times.
Haut & Barch
(2006)
49 26 23 Gender Episodic
encoding, 2-back
and yes/no
recognition tasks
Brain activations
by fMRI,
response time and
accuracy
ANOVA and t-
test
No sig. gender
difference
Reavis &
Overman
(2001)
161 95 66 Gender Decision tasks
(Iowa Card Task
and California
Weather Task)
Performance ANOVA Males had better
performances
Bolla et al.
(2004)
20 10 10 Gender A decision task
(Iowa Card Task)
Performance,
brain activiations
by PET
Mann Whitney U-
Test
Males had better
performances, males
showed greater
lateralized brain activity
to the right hemisphere,
females showed grater
brain activity in the left
DLPFC
Crone et al.
(2003)
257 Not
stated
Not
stated
Gender A decision task
(Iowa Card Task)
Performance ANOVA No sig. gender
difference
Lighthall et al.
(2012)
47 23 24 Gender A decision task
(Balloon Analogue
Risk Task)
Performance,
brain activiations
by fMRI
ANOVA,
correlation tests
No sig. gender
difference
van der Bos et
al. (2012)
213 140 73 Gender A decision task
(Iowa Card Task)
Deliberation
time, impulsivity
and completion
time
Not stated No sig. gender
difference
20
2.3.2. Personality Differences
Effects of personality on perceived workload are investigated by a few studies
(Friedman & Rosenman, 1974; Grubb et al., 1994; Damos & Bloem, 1985).
Friedman and Rosenman (1974) grouped personality indicators into two main types
as Type A and Type B where people with Type A personality had characteristics of
being impatient, competitive, aggressive, incapable of relaxation and people with
Type B personality had characteristics of being relaxed, patient and easy-going.
They reported Type A participants had higher scores than Type B participants on
tests. Damos & Bloem (1985) found that only one between-group difference was
significant under single-task conditions: Type A participants performed memory
searches almost twice as quickly as Type B participants. Another important finding
of this study is that Type A participants were less satisfied with their performance
although they performed better in some conditions.
Sohn and Jo (2003) examined personality and its effects on mental workload to find
the ideal flight crew combination. They divided their participants into four
personality groups. They found a high relation between personality and NASA-
TLX, which is the overall measure of mental workload.
Rose et al., (2002) used a broader personality trait (the Big Five personality traits) to
examine relationship with both vigilance performance and perceived workload. The
Big Five personality traits is a personality inventory that involves 181 items for self-
reporting and generates scores for each of the five dimensions (McCrae & Costa,
1997). These dimensions are Extraversion, Agreeableness, Conscientiousness,
Emotional Instability (Neuroticism) and Openness to Experience. The results of
Rose et al., (2002) study which investigated influence of the Big Five personality
traits on performance and perceived workload, showed that two of five dimensions
(extraversion and conscientious) correlated with performance. With regard to
perceived workload, high neuroticism associated with frustration levels.
21
Gencoz and Oncul (2012) adapted the Big Five personality traits to be used within
Turkish culture. Their analysis on Turkish people presented us a new personality
characteristic which was called Negative Valence dimension and added to the five
dimensions in the original the Big Five personality traits. Their study created a Basic
Personality Traits Inventory (BPTI) with six personality dimensions (extraversion,
neuroticism, agreeableness, conscientiousness, openness to experience and negative
valence) identified for the Turkish culture.
Since one of the aims of this thesis study is to investigate whether personality has an
effect on mental workload during different cognitive tasks, using a generally
accepted personality test is important. Since our experiments were planned to be
done in Turkey, Gencoz and Oncul (2012)’s study was very helpful and their BPTI
with six dimensions was used in this study.
There are few studies used personality traits with brain imaging techniques.
DeYoung (2010) used Big Five personality traits to investigate the association of
each trait with different brain areas. As a result, four of five traits (extraversion,
neuroticism, agreeableness and conscientiousness) supported their hypothesis.
Extraversion has a significant positive association with the activations in medial
orbitofrontal cortex. Neuroticism is negatively correlated with the activations in
right dorsomedial PFC and in portions of the left medial temporal lobe, positively
correlated with the activations mid-cingulate cortex. Agreeableness is associated
positively with the activations in the retrosplenial region of posterior cingulate
cortex and fusiform gyrus, associated negatively with the activations in superior
temporal sulcus. Lastly, conscientiousness has a significant positive association with
the activations in lateral PFC and a significant negative association with the
activations in posterior fusiform gyrus.
There are many types of personality traits and different kinds of cognitive tasks and
hence the relationship between personality and cognitive ability is a quite wide
22
research area. However, results of relationship between cognition tasks and
personality are not always consistent. Extraversion, conscientiousness and
agreeableness change according to type of cognitive task. Extraversion was
positively correlated with speed and memory, but it was negatively correlated with
verbal ability (Chamorro-Premuzic et al., 2006). According to Costa et al. (1976),
high anxiety was associated with lower cognitive performance, while high openness
to experience and low extraversion were associated with higher cognitive scores.
Graham & Lachman (2012) found that personality stability was associated with
better cognitive performance. Conscientiousness had both positive and negative
relationships with cognitive tasks (Furnham & Chamorro-Premuzic, 2006; Moutafi
et al., 2006). Results of Graham and Lachman (2014) showed that openness was
positively correlated with verbal fluency, extraversion was negatively correlated
with reasoning, neuroticism was negatively correlated with reasoning. Judge et al.
(1999) claimed that high extraversion, high conscientiousness, low agreeableness
and low neuroticism were associated with extrinsic career success, when high
conscientiousness was associated with intrinsic career success.
Specifically, Davies (1965) used two Heron subscales (which measure emotional
stability and sociability) as a personality trait during his experiment and found that
there was no significant correlation between maze test scores and personality. The
Big Five personality traits has recently been used in many studies that focus on the
effects of personality type on decision-making tasks that involves delay discounting,
reward sensitivity, gambling, and risk-taking since it was first used by Lauriola and
Levin (2001). According to Lauriola and Levin (2001)’s results, people with low
neuroticism and people with high openness to experience took more risk in a
decision making task, while the correlation coefficients were not significant between
risk-taking and other personality traits (extraversion, conscientiousness,
agreeableness). Hooper et al. (2008) found that there was a negative relationship
between neuroticism and performance in a decision-making task. Byrne (2015) used
the Big Five personality traits to examine the relationship between a decision
23
making task and personality and he found that agreeableness, conscientiousness and
neuroticism negatively correlated with performance in the decision making task.
Table 2 summarizes the literature review on personality differences.
24
Table 2: Summary of Studies on Personality Differences
Study N Females Males Personality Test Cognitive Task Measurement Analyses Findings
Friedman
&Rosenmann
(1974)
Not
stated
Not
stated
Not
stated
Type A and Type B Cognitive tests Performance Not stated Type A participants
had higher scores
Damos & Bloem
(1985)
16 16 0 Type A and Type B Single and dual-
tasks
Reaction time
and accuracy
Not stated Type A participants
performed faster
Sohn and Jo
(2003)
61 Not
stated
Not
stated
Myers Briggs Type
Indicator
4 – 6 flight Heart rate,
altitude
deviation, and
NASA-TLX
Two-way
MANOVA, one-
way MANOVA
A high relation
between personality
and NASA-TLX
Rose et al.,
(2002)
96 48 48 The Big Five
personality traits
A vigilance task Performance,
false alarm,
reaction time and
NASA-TLX
One-way ANOVA,
mixed ANOVA,
partial correlation
Extraversion and
conscientious
correlated with
performance,
neuroticism
correlated with
workload
DeYoung (2010) 116 58 58 The Big Five
personality traits
Not stated Brain activations
by fMRI
Correlation and
regression-based
analyses
Brain activations
correlated with
extraversion,
neuroticism,
agreeableness and
conscientiousness
traits
Chamorro-
Premuzic et al.
(2006)
118 87 31 Eysenckian
personality
Verbal and
numerical
cognitive tasks
Performance Correlation and
regression-based
analyses
Verbal ability was
associated with
introversion,
dissimulation and
neuroticism.
Numerical ability was
associated with
caution
Costa et al.
(1976)
969 0 969 Anxiety,
Extraversion and
Openness to
Experience
Information
Processing
Ability (IPA),
Manual Dexterity
(MD), and
Pattern Analysis
Capability (PAC)
Performance Correlation and
regression-based
analyses
High anxiety was
associated with lower
performance in all
tasks, high openness
to experience was
associated with
higher IPA and PAC
scores, low
extraversion was
associated with
higher PAC
Graham &
Lachman (2012)
4974 Not
stated
Not
stated
Personality Stability Cognitive tasks Performance and
reaction time
Multiple regression
and ANCOVA
Personality stability
was associated with
better cognitive
performance
Furnham &
Chamorro-
Premuzic (2006)
93 70 23 The Big Five
personality traits
Examinations, a
cognitive ability
and beliefs about
intelligence test
Examination
grades, cognitive
ability
performance and
beliefs about
intelligence
performance
Partial correlation,
ANOVA
Conscientiousness,
extraversion
significantly
correlated with
examination grades.
There was a
correlation between
conscientiousness and
beliefs about
intelligence
performance.
25
Table 2 (continued): Summary of Studies on Personality Differences
Study N Females Males Personality Test Cognitive Task Measurement Analyses Findings
Moutafi et al.
(2006)
2658 Not
stated
Not
stated
The Big Five
personality traits
Cognitive tasks Fluid intelligence Correlational
analyses
Openness was
positively correlated
with fluid
intelligence, while
conscientiousness
was negatively
correlated with it.
Graham and
Lachman (2014)
154 73 81 The Big Five
personality traits
A category
fluency
task, generating
number, counting
and Stop and Go
task.
processing speed,
reaction
time,verbal
fluency,
reasoning,
memory
Hierarchical
multiple regression
openness positively
correlated with
verbal fluency,
extraversion
negatively correlated
with reasoning,
neuroticism
negatively correlated
with reasoning.
Judge et al.,
1999
354 Not
stated
Not
stated
The Big Five
personality traits
Jon satisfaction,
status, income,
intrinsic success
and extrinsic
success
Correlation and
regression-based
analyses
High extraversion,
high
conscientiousness,
low agreeableness
and low neuroticism
were associated with
extrinsic career
success, when high
conscientiousness
was associated with
intrinsic career
success
Davies (1965) 540 240 300 the Heron Inventory The Perceptual
Maze Test
Performance Product-moment
correlation
No sig. personality
difference
Lauriola and
Levin (2001)
Not
stated
Not
stated
Not
stated
The Big Five
personality traits
the Ambiguity-
Probability
Tradeoff Task
Risk-taking
performance
Correlation and
regression-based
analyses
Low neuroticism and
high openness to
experience were
associated with
greater tisk taking
Hooper et al.
(2008)
Not
stated
Not
stated
Not
stated
Eysenckian
personality
A decision task
(Iowa Card Task)
Performance Multiple regression
analyses
There were a
relationship between
neuroticism and
performance and a
relationship between
extraversion and
performance
Byrne (2015) 127 76 51 The Big Five
personality traits
A dynamic
decision making
task
Completion time
and performance
t-test, correlational
analyses and two-
step hierarchical
multiple regression
Agreeableness,
conscientiousness and
neuroticism
negatively correlated
with performance in
the decision making
task
Gray (2001) 152 76 76 Extraversion and
neuroticism
Spatial or verbal
2-back task
Reaction time
and accuracy
mixed ANOVA Extraversion and
neuroticism did not
have any correlations
with reaction time and
accuracy
Canli et al.
(2001)
Not
stated
Not
stated
Not
stated
The Big Five
personality traits
Emotional
experience
Brain activations
by FMRI
Correlation and
regression-based
analyses
Brain activations
positively correlated
with extraversion and
negatively correlated
with neuroticism.
26
27
CHAPTER 3
THE METHODOLOGY
In this chapter, information about participants and apparatus are given. Cognitive
tasks used in experiments and procedures for experiments are discussed in detail.
Before experiments, all participants signed the Informed Consent Form approved by
the Human Subject Ethics Committee at the Middle East Technical University as
seen in Appendix A.
3.1. Participants
Participation to the study was voluntary. 37 right-handed participants (23 female, 14
male) were recruited for this study. Their mean age was 25.51. They were
undergraduate or graduate engineering students at the Middle East Technical
University or Hacettepe University. Demographical data of participants is shown in
Appendix B.
3.2. Apparatus
This section includes the description of materials, software and systems used during
the experiments. In this study a questionnaire about user information, the Informed
Consent Form, the Basic Personality Traits Inventory, fNIR Device, an n-back task,
a maze task, an information sampling task were used. In the following sections these
apparatus will be explained in detail.
3.2.1. Functional Near-Infrared (fNIR) Spectroscopy
28
fNIR technology is sensitive to mental workload which is provided by measuring the
change in the rate of oxygenated and deoxygenated hemoglobin in the prefrontal
cortex. Continuous wave NIR spectroscopy system supplied by fNIR Devices LLC
(Ptomac, MD: www.fnirdevices.com) was used during all experiments. The system
included a control box, two sensor pads (each piece contains 2 channels) and a
COBI Control Device Software (Figure 1). These two sensor pads were designed to
monitor dorsal and inferior frontal cortical areas of the brain (Ayaz et al., 2012), so
they were set up on right and left area of the forehead respectively (Figure 2). Data
acquisition, collection and presentation were provided by Cognitive Optical Brain
Imaging Studio (COBI) Control Device and fNIRSoft Software (Ayaz et al., 2012).
COBI is a software where users are able to acquire, process and visualize fNIR
signals, whereas fNIRSoft offers tools to filter and analyze raw fNIR signals.
Figure 1: fNIR System Design
29
Figure 2 - fNIR Sensors Placed on the Participant’s Forehead
On each sensor pad, there is a light source which emits light changes between 700-
900 nm wavelengths. The light in this frequency band is primarily absorbed by
hemoglobin molecules, so changes in the concentration of oxy-hemoglobin and
deoxy-hemoglobin can be observed by the sensors through the channels (Ozcan,
2012). After measuring hemodynamic changes, analyses were performed by using
fNIRSoft software. The increase in oxy-hemoglobin concentration with respect to
deoxy-hemoglobin concentration indicates brain activation and functional
challenges (Izzetoglu, 2004). Therefore, fNIR will be helpful for measuring mental
workload.
3.2.2. Information Sampling Task
The Information Sampling Task (IST) was used as a decision making cognitive task.
The IST was presented for the first time by Clark et al. (2006).
The IST is a measure to gather information about people’s tendency to make a
decision. The task involves a five-by-five matrix of closed boxes and opened forms
of these boxes are one of two colours. The aim of the task is finding which colour is
30
dominant to the other one. Subjects can open these boxes by clicking on them and
they are free to open as many as they want to decide which colour is the majority in
the 5x5 matrix. When subjects decide on the dominant colour in the matrix at some
point, they press on one of the two colours in the screen. The process can be seen in
Figure 3.
There are two levels of the task with 10 trials in each of them: Fixed win (FW) and
decreasing win (DW). In the FW, subjects start with 0 points to the task, they can
open as many boxes as they want and number of opened boxes does not affect their
points. After their decision, if it is correct they win 100 points. Otherwise, they lose
100 points. In the DW level, subjects start with potential 250 points, and every box
opened by subjects decreases the amount of the possible award by 10 points. At the
decision point, if their decision is correct, they win the leftover of 250 points; if it is
not, they lose 100 points. The equations used to calculate for IST-1 and IST-2
performances are as follows, respectively:
𝐴1= ∑ 100 𝐷𝑖10𝑖=1 , Eq.1
𝐴2= ∑ ((250 − 𝑛𝑖) 𝐾𝑖10𝑖=1 − 100(1 − 𝐾𝑖)), Eq.2
where Di is equal to 1 if the decision of the participant is correct at trial-i during
IST-1 and Di is equal to -1 if it is not. For IST-2, Ki is used for accuracy. Ki is equal
to 1 if the decision of the participant is correct at trial-i during IST-2 and Ki is equal
to 0 if it is not.
The script for the IST was supplied by Millisecond Software LLC
(www.millisecond.com) and it was run on a computer during the experiment.
31
Clark et al. (2006) used trial numbers (numbers of boxes opened), errors, latency of
box opening and total points for his analyses and results. In this study, trial numbers,
completion time of the task and performance were used for analyses.
Figure 3: Screen Display for the IST (Clark et al., 2006)
3.2.3. N-Back Task
N-back task is a standard attention and working memory task (Watter, 2001). 3x3
matrix was shown with an object which was occupied in any position in matrix. The
stimuli duration was 500 milliseconds and there were 2500 milliseconds until the
next stimuli. At each stimulus, the object could be at any place in the matrix. In this
study, four conditions were used from one to four to create different difficulty
levels. As each stimulus was presented, the participant had to compare the position
of the current stimulus with the stimulus that occurred n items before. The
32
participant was asked to keep track of the position of the object and in the 1-back
condition, if the position of the object was identical to the one presented previously,
the participant had to press ‘A’ button. In the 2-back, 3-back and 4-back condition,
if the object’s position was identical to the one presented two, three and four trials
back, respectively, the participant had to press ‘A’ button. Each condition had 20
trials. A screen display of 2-back condition can be seen in Figure 4.
Figure 4: Screen Display for N-back Task
Brain Workshop 4.8 was used for n-back task which was supplied from Brain
Workshop web page (http://brainworkshop.sourceforge.net/). The program gave
accuracy results of each n-back. N-back accuracy was calculated by the following
equation:
33
𝐴3 =𝑀
𝑃 +𝑈 , Eq.3
where M is the number of pressing the button when position is matched, U is the
number of pressing the button when position is unmatched, and P is the number of
total same positions of objects.
The completion time was the same for every participant, because stimuli had same
trials in each experiment. Therefore, only accuracy of each task was used as a
performance measure.
3.2.4. Maze Task
Maze is a task that demands visual-spatial skills (Ayaz et al., 2008). Maze contains
both working memory function to remember where you are and decision-making
task to decide turning to right or left. Spatial ability is a very common study area on
gender differences. Therefore, maze is chosen as another cognitive task to be used in
our study.
Maze Suite application was used as a maze task. A 3-D maze environment was
created by using Maze Suite which was first described and presented by Ayaz et al.
(2008), copyrighted by Drexel University and obtained from Maze Suite webpage
(mazesuite.com/downloads). It is an application to create 3D virtual environments. It
is used for researches based on navigational and spatial cognitive neuroscience
experiments. Maze Suite is composed of three applications: MazeMaker,
MazeWalker and MazeAnalyzer. MazeMaker is the editor application to create and
edit experiments. MazeWalker is the application that renders mazes created by
MazeMaker. Lastly, MazeAnalyzer is the application for analysis.
Having full control of designing, running and analyzing in one application is a
benefit for our spatial based experiment. Another advantage of Maze Suite is its
34
ability to send signals to fNIR through a computer serial port which is an advantage
for synchronizing fNIR measures with behavioral performance data.
3.2.4.1. Maze Maker
Maze Maker is the first phase of Maze Suite application. 3D maze environments are
designed in this step.
The path, start and finish points, any different objects on the path were created in
there. The maze should not have been too easy since the focus of this study is mental
workload. But also, we did not want to design a very complicated maze that people
would struggle to finish very hardly and maybe give up by considering all other
cognitive tasks. The top view of the maze design used in this study is shown in
Figure 5.
Figure 5: Design of the Maze in Maze Walker
35
3.2.4.2. Maze Walker
Maze Walker is the enabler of Maze Suite (Ayaz et al, 2008). The maze that will be
used is chosen and started with Maze Walker. An initialization dialog box of this
application provides for arranging initial settings of the maze such as video
(resolution, colour depth, full screen), control (use mouse/joystick’ use pre-recorded
log file), visual (show crosshair/timer/bar, enable lights/shaders, skip warnings),
maze (selection of maze) and logging (automatically save a log file, enter the name
of user) (Figure 6). After all settings are done, users are able to start the maze by
clicking on “Start” on this application.
36
Figure 6: Initial Setting Screen of Maze Walker
37
In this study, resolution was set up 800x600, colour depth was chosen highest
resolution (32 Bits) and full screen option was checked to provide full concentration.
Keyboard use during maze was requested from users and mouse use was forbidden
to ensure equal condition. The maze that was opened was the one whose design is
shown in Figure 6. Auto log was checked to ensure all log files would be recorded
because they would be used in analysis phase.
3.2.4.3. Maze Analyzer
The analysis phase of Maze Suite is Maze Analyzer. Log files that are recorded
during maze task can be opened and user behavior can be investigated. Completion
time of the maze, length of the path that is toured until finish line and visual of the
path that is travelled in the maze map can be given by Maze Analyzer after the task
is completed as seen in Figure 7. In this study, completion time and path length were
used as performance measures.
Figure 7: Maze Analyzer
38
3.2.5. Basic Personality Traits Inventory
After three cognitive tasks were completed, participants were asked to complete the
Basic Personality Traits Inventory (BPTI). BPTI with six dimensions which is
validated for use with the Turkish population (Gencoz and Oncul, 2012) was used as
a personality trait and 45 items were ranked by participants. The questionnaire and
English version of the questionnaire can be seen in Appendix C and D, respectively.
This was the last step of the procedure.
These 45 items gave results under six dimensions which were extraversion,
conscientiousness, agreeableness, neuroticism, openness to experience and negative
valence. At each item, participant was asked to rate his/her familiarity with the
adjective from 1 to 5. There were 8 items for extraversion, 8 for conscientiousness, 8
for agreeableness, 9 for neuroticism, 6 for openness to experience and 6 for negative
valence dimensions respectively. Arithmetic means for each dimension was
calculated by using items belonging themselves. Some of items are reversing entry,
so their point was taken as six minus their actual point. Then, arithmetic mean
calculation was done. Which items belong to which dimension is shown in
Appendix E. The results of the inventory gave an idea about how the person shows
similarity with the dimension in a scale of 1 to 5. In this study, performance and
fNIR measures were analyzed for each dimension.
3.3. Experimental Procedure
The experiments were conducted at the METU Informatics Institute. Experimental
setup included an fNIR system and a computer to run all the tasks.
The aim of this study, its procedure and tasks were explained to all of the subjects
orally. User information questionnaire was applied before the experiment to obtain
demographics data from the participants, which can be seen in Appendix B. Then,
the participants read and signed the Informed Consent Form (Appendix F). Later,
39
sensor pads of fNIR system were placed on the participants’ forehead, signal quality
check was performed and baseline measures were obtained.
SAN Suite software was used for automated execution of the experimental protocol.
The break between the tasks were 15-20 seconds. After their comfort was ensured,
the other phase of the experiment was started. The experiment had three phases (n-
back, IST, maze) and they were chosen randomly to prevent sorting effects.
During all three tasks, fNIR pads were on the participants’ forehead and were not
removed. After completing each task, participants were asked if the pads were
giving them any discomfort, and if so the fNIR pads were removed for a short
duration to give them some relief. Participants were informed that they can quit the
experiment at any time they wish due to any discomfort they may experience.
However, nobody felt that way and there were not any incomplete experiment.
Following the completion of the cognitive tasks, the Turkish culture based basic
personality traits inventory based on 45 items was completed by participants.
N-back task took 5-6 minutes in length with a break between levels. The total time
of the IST changed from person to person, but it was typically between 7 and 15
minutes. Maze task lasted between 1 and 6 minutes. Overall, completion of the
experiment took 45 minutes on average, including the initial introduction, all tasks
and the questionnaires.
40
41
CHAPTER 4
RESULTS AND DISCUSSIONS
In this chapter, results of analyses are given and discussed.
4.1. Data Analysis
Data of 37 participants (23 females and 14 males) were included for the analysis of
performance measures. However, due to excessive noise and motion artifacts some
of the subjects had to be excluded from the fNIR analysis. In particular, data from
19 participants (8 females and 11 males) could be used for the fNIR analysis.
There are 2 detectors on each fNIR sensor and each of them provides data on
changes in the hemoglobin and oxy-hemoglobin levels. The raw fNIR Data (4
channels x 2 wavelengths) were filtered by FIR (Finite Impulse Response Digital
Filter) to decrease high frequency noise, respiration and cardiac cycle effects (Ayaz
et al., 2012). An approach SMAR (Sliding- Window Motion Artifact Rejection)
(Ayaz et al., 2010) was performed for motion artifact detection and rejection from
the refined fNIR measures. fNIRSoft was used to filter the data and calculate the
blood oxygenation levels. fNIR measures data of 19 participants out of 37
participants were available to be used after filtering data by FIR and SMAR to
eliminate high frequency noise, respiration, cardiac cycle effects, saturation and
head motion artifacts. Since activations in voxel-1 and voxel-4 were enough for
cognitive tasks, we decided to use them.
42
This study used maze completion time, maze path length, IST-1 performance, IST-1
trial number, IST-1 completion time, IST-2 performance, IST-2 trial number, IST-2
completion time, 1-back performance, 2-back performance, 3-back performance,
and 4-back performance as performance measures. The study also used maze
oxygenation levels in voxel-1 (left DLPFC) and voxel-4 (right DLPFC); IST-1 and
IST-2 oxygenation levels in voxel-1 (left DLPFC) and voxel-4 (right DLPFC); 1-
back, 2-back, 3-back, and 4-back oxygenation levels in voxel-1 (left DLPFC) and
voxel-4 (right DLPFC) as fNIR measures.
SPSS 20 for Windows (SPSS Inc., Chicago,IL, USA) was used for data analysis.
First, boxplots and histograms of variables were drawn for descriptive analysis of
collected data. There were outliers on maze completion time, maze path length, IST-
1 completion time, IST-2 completion time, IST-2 trial number. First, reasons of
these outliers were investigated. However, all data were entered correctly and
participants did not have any unexpected situations. There was not any problem in
collecting data, so eliminating outliers was out of option and transformations of
these outliers were considered. Since maze completion time, maze path length, IST-
1 and IST-2 completion time show positive skewness; square root transformations
were applied (Field, 2009). However, data transformations were not able to
eliminate the outliers. Only one solution was left for outliers: changing the score.
The mean plus two times the standard deviation was applied. There was also an
outlier in trial number on IST-2 data and it did not suit any form of transformation
rule. Since there was not any special case and deleting the value seemed losing a
data without a reason, changing the score was decided. The mean plus two times the
standard deviation was also applied for that data set (Field, 2009). After these
changes, new boxplots of maze and IST performance measures were drawn. All
boxplots of performance and fNIR measures are shown in the appendices. Boxplots
for maze completion time, maze path length, IST performances, IST completion
times, IST trial numbers, n-back performances can be seen in Appendix G, H, I, J, K
and L, respectively. fNIR measures during Maze task, ISTs and n-back tasks are
43
given in Appendix M, N and O, respectively.
All data were examined for normality by using the Shapiro-Wilk test. While all
fNIR data are normally distributed (Table 3), some performance variables violated
the assumptions of normality (Table 4). In Table 3, Shapiro-Wilk tests for all
variables are non-significant (p > .05) which shows that these samples are not
significantly different from a normal distribution. However, Table 4 shows that the
distributions of maze path length, maze time, IST-1 performance, IST-2 completion
time, 1-back performance and 2-back performance are significantly different from a
normal distribution (p < .05). When the assumptions of normality are not satisfied,
parametric tests cannot be used. Therefore, nonparametric tests were applied for
these variables. The flowchart of data analysis steps is given in Figure 8.
Table 3: Normality Tests for fNIR Variables
Statistic Df Sig. Statistic df Sig.
V1_maze 0.133 19 .200* 0.939 19 0.254
V4_maze 0.131 19 .200* 0.977 19 0.897
V1_ist1 0.101 19 .200* 0.973 19 0.829
V4_ist1 0.136 19 .200* 0.966 19 0.694
V1_ist2 0.151 19 .200* 0.957 19 0.517
V4_ist2 0.117 19 .200* 0.982 19 0.966
V1_1back 0.123 19 .200* 0.976 19 0.891
V4_1back 0.15 19 .200* 0.964 19 0.644
V1_2back 0.124 19 .200* 0.973 19 0.834
V4_2back 0.113 19 .200* 0.982 19 0.962
V1_3back 0.113 19 .200* 0.981 19 0.952
V4_3back 0.09 19 .200* 0.966 19 0.686
V1_4back 0.09 19 .200* 0.98 19 0.944
V4_4back 0.138 19 .200* 0.97 19 0.773
Kolmogorov-Smirnov Shapiro-Wilk
*. This is a lower bound of the true significance.
44
Table 4: Normality Tests for Performance Variables
Figure 8: Flowchart of Data Analysis
Statistic Df Sig. Statistic df Sig.
MazePath 0.147 37 0.043 0.94 37 0.047
MazeTime 0.114 37 .200* 0.938 37 0.041
IST1Per 0.299 37 0 0.791 37 0
IST1Time 0.109 37 .200* 0.95 37 0.094
IST1TrialNo 0.09 37 .200* 0.953 37 0.118
IST2Per 0.095 37 .200* 0.984 37 0.869
IST2Time 0.195 37 0.001 0.898 37 0.003
IST2TrialNo 0.1 37 .200* 0.957 37 0.157
back1 0.43 37 0 0.638 37 0
back2 0.155 37 0.024 0.898 37 0.003
back3 0.083 37 .200* 0.978 37 0.66
back4 0.082 37 .200* 0.968 37 0.367
Kolmogorov-Smirnov Shapiro-Wilk
*. This is a lower bound of the true significance.
Applying Appropriate Tests
Examining Normality
Changing the Score for Eliminating Outliers
Trying Data Transformations
Checking for Outliers
Filtering the Data
Data Collection
45
4.2. Gender Difference in Performance Measures
Mann Whitney U-Tests were conducted for 37 participants (23 females and 14
males) to see whether there was gender difference in performance during maze task
for each maze performance measure: maze completion time and maze completion
length. Results showed that the mean maze completion time did not significantly
differ between males (M=188.79, SD=91.855) and females (M=255.45,
SD=116.510) where U=104.5, z=-1.770, p>.05. Figure 9 shows average values for
males and females in maze completion time. Summary of the test can be found in
Table 5.
Figure 9: Boxplot of Maze Completion Time according to Gender
46
For another maze performance measure, results showed that maze path length did
not significantly differ between males (M=195.75, SD=125.960) and females
(M=193.20, SD=125.110) where U=156, z=-0.157, p>.05. Figure 10 shows average
female and male values in maze path length. Summary of the test can be found in
Table 5.
Figure 10: Boxplot of Maze Path according to Gender
47
Table 5: Mann Whitney U-Test Summary for Maze Performance Measures
As a result it can be concluded that there is not a significant gender difference in
maze performance measures. There are many studies about gender differences on
spatial ability and the results of these studies range from males outperforming
females (Galea & Kimura, 1993; Moffat, 1998; Waller et al., 2001) to no gender
differences (O’Laughlin & Brubaker, 1998; Taylor & Tversky, 1992; Sadalla &
Montello, 1989; Montello & Pick, 1993).
When we considered only maze-based tasks and excluded other kinds of spatial
ability tests, the studies (Galea & Kimura, 1993; Persson, 2013; Moffat, 1998
indicate males are significantly better than females on maze tasks. However, we
found that performances of females and males are not significantly different from
each other in our study. There may be two reasons for this result. First reason of this
mismatch may be based on the demographic properties of the participants. The
participants of previous studies were students who enrolled in a psychology course,
whereas the participants of our study were undergraduate or graduate students in an
engineering degree program. People in our study were more likely to have spatial
perception skills since spatial-visual skills are essential for success in engineering.
Some of the engineering courses may increase students’ ability on spatial and visual
tasks, especially design based courses; therefore engineering students may have
Maze
Completion Time
Maze Path
Length
Mann-Whitney U 104.500 156.000
Wilcoxon W 209.500 432.000
Z -1.770 -.157
Asymp. Sig. (2-tailed) .077 .876
Exact Sig. [2*(1-tailed Sig.)] 0.077 0.889
48
higher spatial ability. That fact may be the reason for why there was no a gender
difference in the maze task.
Second, the maze task in our study was an easier and shorter than the maze tasks
used in other studies (Moffat, 1998; Waller, 2001). In our study there were other
cognitive tasks to be completed by the participants, so we opted for shorter maze
designs. In our study there was one trial that lasted between 2 to 7 minutes. Moffat
(1998) used 4 practice trials and 5 trials on each of two experimental mazes.
Similarly, a two-hour long experiment was employed in Waller (2001)’s study.
Therefore, using longer maze tasks or increasing the number of trials may underlie
the gender difference reported in those studies. A reorganized maze according to
these conditions can be applied for future studies.
For the information sampling task, IST-1 performance and IST-2 completion time
did not satisfy the assumptions of normality as shown before, therefore a Mann
Whitney U-Test (a nonparametric test) was conducted for 37 participants (23
females and 14 males) to examine whether there was gender difference in
completion time / performance during IST-1 Task / IST-2 Task. Results showed that
the mean IST-1 completion time did not significantly differ between males
(M=184.44, SD=64.92) and females (M=146.19, SD=55.89) where U=109.5, z=-
1.613, p>.05. For the second phase of the IST, the mean IST-2 completion time did
not significantly differ between males (M=184.55, SD=62.91) and females
(M=146.38, SD=70.88) where U=102.5, z=-1.832, p>.05. Figure 11 shows boxplots
for males and females in IST-1 and IST-2 completion time. Summary of the test can
be found in Table 6.
49
Figure 11: Boxplot of IST Completion Times according to Gender
Briefly, results show that there is no gender difference in completion time during
both information sampling tasks. This finding is compatible with the literature
(Crone et al., 2003; Lighthall et al., 2012). Completion time of a decision-making
task did not change according to gender in previous studies.
A Mann Whitney U-Test was applied for 37 participants (23 females and 14 males)
to investigate the gender difference in performance of IST-1 and IST-2 tasks. Result
showed that the mean IST-1 performance did not significantly differ between males
(M=900.00 SD=188.11) and females (M=765.22, SD=205.84) where U=100, z=-
2.055, p>.05. Similarly, for the second phase of the IST, the mean IST-2
performance did not significantly differ between males (M=376.00, SD=114.37) and
females (M=414.09, SD=176.56) where U=136.5, z=-0.767, p>.05. Figure 12 shows
average values for males and females in IST-1 and IST-2 performance. Summary of
50
the test can be found in Table 6.
Figure 12: Boxplot of IST Performances according to Gender
Previous studies involve two kinds of results: males outperformed females in
decision-making tasks (Reavis & Overman, 2001; Bolla et al., 2004) and no gender
differences during a decision-making task (Crone et al., 2003; Lighthall et al., 2012).
The results of this study show that there is no gender difference in completion time
during both information sampling tasks. The participants of this study were
engineering students at well-known universities which means they performed
similarly at the college admissions test and they experienced a similar training in
engineering. Therefore, their cognitive abilities are possibly very close to each other
51
which may be the reason od observing no gender difference in this sample.
Table 6: Mann-Whitney U-Test Summary for IST Performance Measures
Since IST-1 and IST-2 trial number satisfied the assumptions for a parametric test,
independent t-tests were run for 37 participants (23 females and 14 males) to check
for gender differences in both trial numbers at IST-1 and IST-2. For the first phase
of the IST, results showed that trial number did not significantly differ between
males (M=189.50 SD=57.51) and females (M=156.39, SD=48.33) where
t(35)=1.881, p>.05. However, results showed that IST-2 trial number significantly
differed between males (M=152.43 SD=39.35) and females (M=122.78, SD=40.77)
where t(35)=2.173, p<.05. Trial numbers of males were significantly higher than
trial numbers of females. Figure 13 shows average values for males and females in
IST-1 and IST-2 trial numbers. Summary of the test can be found in Table 7.
Results show that males perform significantly more trials than females during IST-2,
even though there is no gender difference in the number of trials during IST-1. Since
there is no reward/loss for opening boxes at IST-1, behavior difference in gender is
not expected and the result supports this expectation. However, males open more
boxes during IST-2 where each opening decreases the reward. Previous studies (van
den Bos et al. 2013; Overman et al., 2006; Stoltenberg & Vandeder, 2010) show that
IST1- Time IST2-Time
IST1 -
Performance
IST2 -
Performance
Mann-Whitney U 109.500 102.500 100.000 136.500
Wilcoxon W 385.500 378.500 376.000 241.500
Z -1.613 -1.832 -2.055 -.767
Asymp. Sig. (2-tailed) .107 .067 .040 .443
Exact Sig. [2*(1-tailed Sig.)] 0.107 0.066 0.057 0.448
52
males chose long-term pay-off cases, whereas females prefer larger but infrequent
reward choices during reward oriented, investment-based tasks. Females tend to
make their choices according to their instincts, whereas males go for more facts and
data after analyzing the situation (Agor, 1986; Parik et al., 1994). Thus, that shows
how our finding about trial numbers was compatible with literature.
Figure 13: Boxplot of IST Trial Numbers according to Gender
53
Table 7: Independent t-Test Summary for IST Trial Numbers
Since 1-back and 2-back failed the assumptions of normality, a Mann Whitney U-
test which is a nonparametric test was conducted for 37 participants (23 females and
14 males) to test for gender difference at each task. Results showed that there was no
significant difference between males (M=91.79, SD=14.214) and females (M=92.09,
SD=13.031) during the 1-back task where U=160.5, z=-0.19, p>.05. There was no
significant difference between males (M=78.86, SD=17.20) and females (M=76.65,
SD=23.348) during the 2-back task as well where U=158, z=-0.95, p>.05. Summary
of these tests can be found in Table 8. T-tests were conducted for 3-back and 4-back
tasks. Results showed that there was no significant difference between males
(M=43.36, SD=19.956) and females (M=42.09, SD=19.660) during 3-back task
where F=.035, p>.05. However, males 4-back performance (M=41.21, SD=13.735)
significantly differed from 4-back females performance (M=26.13, SD=16.926)
Lower Upper
Equal
variances
assumed
1.681 .203 1.881 35 .068 33.109 17.604 -2.628 68.846
Equal
variances
not
assumed
1.801 23.963 .084 33.109 18.380 -4.828 71.046
Equal
variances
assumed
.080 .779 2.173 35 .037 29.646 13.642 1.951 57.340
Equal
variances
not
assumed
2.192 28.379 .037 29.646 13.522 1.965 57.327
Std. Error
Difference
95%
Confidence
Interval of the
Difference
IST1Trial
No
IST2Trial
No
Levene's Test
for Equality of
Variances t-test for Equality of Means
F Sig. t df
Sig. (2-
tailed)
Mean
Difference
54
where F=7.916, p<.05. Figure 14 shows average females and males performance for
each back task. Summary of these tests can be found in Table 9.
Table 8: Mann Whitney U-Test Summary for 1-back and 2-back
Table 9: t-Test Summary for 3-back and 4-back
back1 back2
Mann-Whitney U 160.500 158.000
Wilcoxon W 436.500 263.000
Z -.019 -.095
Asymp. Sig. (2-tailed) .985 .924
Exact Sig. [2*(1-tailed Sig.)] 0.988 0.938
Back-3Sum of
Squaresdf
Mean
SquareF Sig.
Between Groups 14.041 1 14.041 0.035 0.852
Within Groups 13901 35 397.173
Total 13915.1 36
Back-4Sum of
Squaresdf
Mean
SquareF Sig.
Between Groups 1980.06 1 1980.06 7.916 0.008
Within Groups 8754.97 35 250.142
Total 10735 36
55
Figure 14: Boxplot of N-back Performances according to Gender
Results suggest that there was no significant difference between gender groups in
terms of their performance during 1-back, 2-back and 3-back tests respectively. This
finding is consistent with the literature (Li et al., 2010; Koch et al., 2007; Schmidt et
al., 2009). However, males are significantly better than females in the 4-back task.
Even though there is not a specific study looking for gender differences in 4-back,
Voyer et al. (1995) show that males had a better performance than females for
spatial-based working memory tasks. Our results for the 4-back trial parallel Voyer
et al.’s finding.
4.3. Task Level Differences in Performance Measures
Maze task consisted of one task while the IST had two phases and the n-back task
had 4 phases, therefore level differences were examined between ISTs and n-back
tasks.
To investigate performance differences between IST-1 and IST-2 tasks, Wilcoxon
Signed Ranks Test was applied for 37 participants (23 females and 14 males) and
56
the results showed that there was a significant difference between performance
scores for IST-1 (M=816.22, SD=207.52) and IST-2 (M=399.68, SD=135.57) tasks
where z=-5.092, p<.001. Figure 15 shows the boxplot for IST-1 and IST-2
performance scores. Summary of these tests can be found in Table 10.
Figure 15: Boxplot of IST Performances according to Task Difficulty
57
Table 10: Wilcoxon Signed Ranks Test Summary for IST Performances
Studies on level difference of IST (Solowji et al., 2012; Townsheed et al. 2006)
found out performance in IST-1 was higher than performance in IST-2 and that
shows how our finding about level difference was consistent with literature.
Wilcoxon Test was applied for 37 participants (23 females and 14 males) to see
whether there was a difference between IST-1 and IST-2 completion time and
results showed that there was no significant difference between completion time of
IST-1 (M=160.66, SD=61.52) and IST-2 (M=168.92, SD=53.73) tasks, where z=-
1.275, p<.05. Figure 16 shows the boxplots of IST-1 and IST-2 completion times.
Summary of these tests can be found in Table 11. There was no significant
difference in completion time and that finding was similar with literature (Solowji et
al., 2012; Townsheed et al. 2006).
IST2_Peformance - IST1_Performance
Z -5.092
Asymp. Sig. (2-tailed) .000
58
Figure 16: Boxplot of IST Completion Times according to Task Difficulty
Table 11: Wilcoxon Signed Ranks Test Summary for IST Completion Times
Paired t-test was conducted for 37 participants (23 females and 14 males) to see
whether there was a significant difference between IST-1 and IST-2 tasks in terms
of number of trials. Results showed that IST-1 trial number (M=168.92, SD=53.730)
was significantly higher than IST-2 trial number (M=134.00, SD=42.274) where
t(36)=6.06, p<.05. Figure 17 shows average IST-1 and IST-2 trial numbers in a
graph. Summary of these tests can be found in Table 12.
IST2_Time - IST1_Time
Z -1.275
Asymp. Sig. (2-tailed) .202
59
Figure 17: Boxplot of IST Trial Numbers according to Task Difficulty
Table 12: Paired t-Test Summary for IST-1 and IST-2 Trial Numbers
Trial numbers during IST-1 was significantly higher than trial numbers during IST-2
and that result was consistent with literature (Clark et al., 2006; Solowji et al.,
2012). Since there is no reward/loss for trials to open boxes during IST-1,
participants opened boxes as many as they wanted. However, penalties for trials
prevent increasing numbers of trials at IST-2.
Lower Upper
IST1TrialNo -
IST2TrialNo34.919 35.059 5.764 23.230 46.608 6.058 36 .000
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviation
Std.
Error
Mean
95% Confidence
Interval of the
Difference
60
N-back task had four levels and ANOVA with repeated measures was conducted for
37 participants (23 females and 14 males) to see whether performance levels of
subjects change according to difficulty level of n-back task. Since F(2.721, 95.234)
= 96.851, p<.001; there was a significant main effect of task level for performance.
There was at least one level differing from another level significantly. Summary of
the test can be found in Table 13. Figure 18 shows average performances for each
level of n-back in a graph.
Existence of performance differences according to task level was also reported in the
literature (Ayaz et al., 2012; Harvey et al., 2005; Blokland et al., 2005). Increasing
complexity of n-back created a significant decrease in n-back performance in
previous studies.
Figure 18: N-back Performance according to Task Difficulty
61
Table 13: Repeated ANOVA Test Summary for N-back Performances
Pairwise comparisons were applied for 37 participants (23 females and 14 males) to
see whether there was a significant difference between any particular pairs of levels
and that brings six research questions as follows:
Do performance levels of subjects change from 1-back to 2-back?
Do performance levels of subjects change from 1-back to 3-back?
Do performance levels of subjects change from 1-back to 4-back?
Do performance levels of subjects change from 2-back to 3-back?
Do performance levels of subjects change from 2-back to 4-back?
Do performance levels of subjects change from 3-back to 4-back?
Results show that subjects’ performance on 1-back (M=91.97, SD=13.30) is
significantly different from 2-back performance (M=77.49, SD=21.00) where p<.05.
Type III
Sum of
Squares df
Mean
Square F Sig.
Sphericity Assumed 80676.545 3 26892.182 96.851 .000
Greenhouse-Geisser 80676.545 2.721 29649.782 96.851 .000
Huynh-Feldt 80676.545 3.000 26892.182 96.851 .000
Lower-bound 80676.545 1.000 80676.545 96.851 .000
Sphericity Assumed 1311.951 3 437.317 1.575 .200
Greenhouse-Geisser 1311.951 2.721 482.161 1.575 .204
Huynh-Feldt 1311.951 3.000 437.317 1.575 .200
Lower-bound 1311.951 1.000 1311.951 1.575 .218
Sphericity Assumed 29154.793 105 277.665
Greenhouse-Geisser 29154.793 95.234 306.137
Huynh-Feldt 29154.793 105.000 277.665
Lower-bound 29154.793 35.000 832.994
Source
back
back *
Gender
Error(back)
62
1-back performance is also significantly different from 3-back performance
(M=42.57, SD=19.66) and 4-back performance (M=31.84, SD=17.27) where
p<.001 for both of them. When we compare 2-back performance with others, it is
seen that 2-back performance (M=77.49, SD=21.00) is significantly different from
3-back performance (M=42.57, SD=19.66) and 4-back performance (M=31.84,
SD=17.27) where p<.001 for both of them. Lastly, the result show that performance
of subjects on 3-back is significantly different from performance on 4-back where
p<.05. Statistics of can be seen from Table 14.
Table 14: Pairwise Comparisons for N-back Performances
Lower
Bound
Upper
Bound
2-back 14,182* 3.555 .002 4.240 24.124
3-back 49,214* 4.153 .000 37.601 60.828
4-back 58,264* 3.173 .000 49.391 67.137
1-back -14,182* 3.555 .002 -24.124 -4.240
3-back 35,033* 4.377 .000 22.791 47.274
4-back 44,082* 4.128 .000 32.537 55.628
1-back -49,214* 4.153 .000 -60.828 -37.601
2-back -35,033* 4.377 .000 -47.274 -22.791
4-back 9.050 4.423 .290 -3.320 21.419
1-back -58,264* 3.173 .000 -67.137 -49.391
2-back -44,082* 4.128 .000 -55.628 -32.537
3-back -9.050 4.423 .290 -21.419 3.320
3-back
4-back
(I) back
Mean
Difference
(I-J)
Std.
Error Sig.
95% Confidence
Interval for
Difference
1-back
2-back
63
4.4. Gender Differences in fNIR Measures
There were 37 participants, but due to excessive noise and motion artifacts, some of
the subjects had to be excluded from the fNIR analysis. 19 participants’ fNIR
measures data out of 37 participants were available to be used after filtering the data.
Since fNIR measures satisfy parametric assumptions, parametric tests were
conducted to see whether there were differences between gender groups. An
independent t-test for the maze task, repeated measures ANOVA for 19 participants
(8 females and 11 males) for IST and n-back tasks were conducted to investigate
gender differences in oxygenation levels observed in the left and right DLPFC
regions. ANOVA for repeated measures for 19 participants (8 females and 11 males)
for IST and n-back tasks were conducted to research gender difference for
oxygenation level in left and right DLPFC regions. The tests were performed for
both oxygenation level in voxel-1 (left DLPFC) and voxel-4 (right DLPFC).
For the maze task, results showed that the mean oxygenation level for left DLPFC
did not significantly differ between males (M=0.752, SD=1.118) and females
(M=0.384, SD=0.937) where p>.05. The mean oxygenation level for right DLPFC
did not significantly differ between males (M=0.220, SD=0.0771) and females
(M=0.778, SD=0.966) where p>.05. Descriptive statistics are presented in Table 15.
Summary of the test can be found in Table 16.
Also t-tests were conducted to see any oxygenation level differences between left
and right DLPFC regions for only males, but results showed that there is no
differences between voxels since p>.05. The same test was applied for only females,
too. But no oxygenation level differences was observed since p>.05.
64
Table 15: Descriptive Statistics for Oxygenation Levels during Maze Tasks
Table 16: Independent Samples Test for Maze fNIR Measures
Literature (Gron et al., 2000) presents that males showed more activation in their left
area of their brain whereas females had more activations than males in their right
prefrontal cortex during a maze task. However, in this study no gender differences
were observed. The fNIR results for gender difference show similarity with
performance results of gender difference for maze task, there is no gender
differences in oxygenation levels.
N MeanStd.
Deviation
Std.
Error
Mean
male 11 0.752 1.118 0.337
female 8 0.384 0.937 0.331
male 11 0.22 0.771 0.233
female 8 0.778 0.966 0.341
Gender
V1_maze
V4_maze
Lower Upper
Equal
variances
assumed
.183 .674 .671 19 .511 .3148 .4689 -.6703 1.2999
Equal
variances not
assumed
.692 16.617 .499 .3148 .4550 -.6469 1.2765
Equal
variances
assumed
3.001 .100 -1.508 19 .149 -.5739 .3805 -1.3733 .2256
Equal
variances not
assumed
-1.426 12.313 .179 -.5739 .4024 -1.4482 .3004
95% Confidence
Interval of the
Difference
V1_maze
V4_maze
Levene's Test t-test for Equality of Means
F Sig. t df
Sig. (2-
tailed)
Mean
Difference
Std. Error
Difference
65
The maze task in our study was easier and shorter than the maze tasks typically used
in other studies, since there are more cognitive tasks in the experiment. Using longer
maze task might have allowed us to see gender differences in brain activation. The
number of participants is 19 and expanding the size of participants can give more
accurate results. Therefore, using longer maze task and larger sampling may be
applied for future studies.
ANOVA for repeated measures was conducted for IST. Results showed that the
mean oxygenation level for left DLPFC did not significantly differ between males
and females where F(1,17)=1964, p>.05. If any significant difference was observed,
a detail analysis would be applied to see at which level a gender difference was
available. However, it was not needed. Descriptive statistics are available in Table
17. Summary of the test can be found in Table 18.
The same test was applied for right DLPFC, too. The mean oxygenation level for
right DLPFC did not significantly differ between males and females where
F(1,17)=200, p>.05. Descriptive statistics are available in Table 17. Summary of the
test can be found in Table 19.
The finding of no gender difference during a decision-making task does not match
with literature. Bolla et al. (2004) investigated brain activity during a decision-
making task and found out that men showed greater lateralized brain activity to the
right hemisphere than females, whereas females showed grater brain activity in the
left DLPFC. The reason of mismatch may be the structure of participants. The
participants of our study for brain imaging were 11 males and 8 females whose age
varied between 21 and 31, while Bolla et al. (2004) used 10 males and 10 females
whose age varied between 21 and 45. Adding people of various ages to participants
might have created diversity which could be the reason of having different results.
66
Table 17: Descriptive Statistics for Oxygenation Levels during ISTs
Table 18: Repeated ANOVA for IST fNIR Measures of Left DLPFC
Table 19: Repeated ANOVA for IST fNIR Measures of Right DLPFC
N Mean Std. Deviation
Std. Error
Mean
male 11 0.451 1.088 0.328
female 8 0.493 0.58 0.205
male 11 0.482 0.884 0.267
female 8 0.634 1.21 0.458
male 11 1.324 0.971 0.293
female 8 0.977 0.891 0.315
male 11 1.309 1.118 0.337
female 8 1.288 0.986 0.349
Gender
V1_ist-1
V4_ist-1
V1_ist-2
V4_ist-2
Effect Value F
Hypothesis
df
Error
df Sig.
Partial Eta
Squared
ist_v1 Pillai's Trace .585 23.949 1 17 .000 .585
Wilks' Lambda .415 23.949 1 17 .000 .585
Hotelling's Trace 1.409 23.949 1 17 .000 .585
Roy's Largest Root 1.409 23.949 1 17 .000 .585
ist_v1 * Gender Pillai's Trace .104 1.964 1 17 .179 .104
Wilks' Lambda .896 1.964 1 17 .179 .104
Hotelling's Trace .116 1.964 1 17 .179 .104
Roy's Largest Root .116 1.964 1 17 .179 .104
Value F
Hypothesis
df
Error
df Sig.
Partial Eta
Squared
Pillai's Trace .462 14.585 1 17 .001 .462
Wilks' Lambda .538 14.585 1 17 .001 .462
Hotelling's Trace .858 14.585 1 17 .001 .462
Roy's Largest Root .858 14.585 1 17 .001 .462
Pillai's Trace .012 0.200 1 17 .661 .012
Wilks' Lambda .988 0.200 1 17 .661 .012
Hotelling's Trace .012 0.200 1 17 .661 .012
Roy's Largest Root .012 0.200 1 17 .661 .012
Effect
ist_v4
ist_v4 * Gender
67
ANOVA for repeated measures was conducted for n-back tasks. Results showed that
the mean oxygenation level for left DLPFC did not significantly differ between
males and females where p>.05. The same test was also conducted for oxygenation
level of right DLPFC during n-back tasks. There was not a significant difference
between males and females since F(3,15)=2936, p>.05. Descriptive statistics are
available in Table 20. Summary of the test can be found in Table 21.
The same test was applied for right DLPFC, too. The mean oxygenation level for
right DLPFC did not significantly differ between males and females where
F(3,15)=1888, p>.05. Descriptive statistics are available in Table 20. Summary of
the test can be found in Table 22.
68
Table 20: Descriptive Statistics for Oxygenation Levels during N-back Tasks
Table 21: Repeated ANOVA for N-back fNIR Measures of Left DLPFC
N MeanStd.
Deviation
Std.
Error
Mean
male 11 0.074 1.288 0.388
female 8 0.061 0.847 0.299
male 11 0.003 0.996 0.3
female 8 0.422 0.73 0.258
male 11 0.981 1.314 0.396
female 8 0.978 1.157 0.409
male 11 0.523 1.364 0.411
female 8 1.22 1.282 0.453
male 11 1.789 1.268 0.382
female 8 1.227 1.476 0.522
male 11 1.215 1.584 0.478
female 8 1.21 1.322 0.467
male 11 2.162 1.992 0.601
female 8 1.182 1.473 0.521
male 11 1.556 1.947 0.587
female 8 1.474 1.34 0.474
V4_3back
V1_4back
V4_4back
Gender
V1_1back
V4_1back
V1_2back
V4_2back
V1_3back
Value F
Hypothesis
df
Error
df Sig.
Partial Eta
Squared
Pillai's Trace .774 17,136 3.000 15.000 .000 .774
Wilks' Lambda .226 17,136 3.000 15.000 .000 .774
Hotelling's Trace 3.427 17,136 3.000 15.000 .000 .774
Roy's Largest Root 3.427 17,136 3.000 15.000 .000 .774
Pillai's Trace .370 2,936 3.000 15.000 .067 .370
Wilks' Lambda .630 2,936 3.000 15.000 .067 .370
Hotelling's Trace .587 2,936 3.000 15.000 .067 .370
Roy's Largest Root .587 2,936 3.000 15.000 .067 .370
Back_v1 * Gender
Effect
Back_v1
69
Table 22: Repeated ANOVA for N-back fNIR Measures of Right DLPFC
fNIR results of n-back task show that there is not a significant gender difference.
Even though there are previous findings which supported gender differences in brain
activation during n-back tasks (Speck et al., 2000; Li et al., 2010), there are also
some studies (Schmidt, 2009; Haut and Barch, 2006) which observed no gender
difference in brain. That variation in the literature creates lack of empirical evidence
for gender difference of working memory in functional brain organization.
4.5. Task Level Differences in fNIR Measures
Izzetoglu et al. (2004) claimed that blood oxygenation level in DLPFC would
increase with increasing task difficulty. However, when task became too difficult, a
break point was reached and the subject did not pay attention to the task anymore.
At that point blood oxygenation level dropped. By considering that hypothesis, a
relationship between changes in blood oxygenation and performance level during a
cognitive task was investigated.
First, their descriptive statistics are shown in Table 23, Table 24 and Table 25
respectively. Oxygenation levels in voxel-1 (Table 23) show us 4-back has the
highest level and 3-back, IST-2, 2-back, maze, IST-1, 1-back are listed from higher
to lower respectively. Oxygenation levels in voxel-4 (Table 24) show us almost
similar results with left DLPFC. 4-back has the highest level and IST-2, 3-back, 2-
Value F
Hypothesis
df
Error
df Sig.
Partial Eta
Squared
Pillai's Trace .563 6,443 3.000 15.000 .005 .563
Wilks' Lambda .437 6,443 3.000 15.000 .005 .563
Hotelling's Trace 1.289 6,443 3.000 15.000 .005 .563
Roy's Largest Root 1.289 6,443 3.000 15.000 .005 .563
Pillai's Trace .274 1,888 3.000 15.000 .175 .274
Wilks' Lambda .726 1,888 3.000 15.000 .175 .274
Hotelling's Trace .378 1,888 3.000 15.000 .175 .274
Roy's Largest Root .378 1,888 3.000 15.000 .175 .274
Effect
Back_v4
Back_v4 * Gender
70
back, IST-1, maze, 1-back are listed from higher to lower respectively.
Table 23: Descriptive Statistics of Oxygenation Levels of Left DLPFC
Table 24: Descriptive Statistics of Oxygenation Levels of Right DLPFC
The average oxygenation levels of left and right DLPC (Table 25) shows that
sequence from higher to lower is as follows: 4-back, 3-back, IST-2, 2-back, maze,
IST-1, 1-back. The boxplots of oxygenation levels for left, right and average are
shown in Figure 19, 20 and 21 respectively.
LEFT MeanStd.
DeviationN
V1_maze 0.5731 1.0124 19
V1_ist1 0.4404 0.8739 19
V1_ist2 1.1893 0.9059 19
V1_1back 0.0688 1.0955 19
V1_2back 0.9795 1.2168 19
V1_3back 1.5522 1.3498 19
V1_4back 1.7490 1.8150 19
RIGHT MeanStd.
DeviationN
V4_maze 0.4338 0.8612 19
V4_ist1 0.5525 0.9786 19
V4_ist2 1.3113 1.0091 19
V4_1back 0.1792 0.8963 19
V4_2back 0.8162 1.3412 19
V4_3back 1.2128 1.4398 19
V4_4back 1.5218 1.6750 19
71
Table 25: Descriptive Statistics of Oxygenation Levels for Average of Left and
Right DLPFC
Figure 19: Boxplot of Oxygenation Levels at Left DLPFC
AVERAGE MeanStd.
DeviationN
Vavg_maze 0.5260 0.7911 19
Vavg_ist1 0.4896 0.8577 19
Vavg_ist2 1.2732 0.9978 19
Vavg_1back 0.1354 0.9231 19
Vavg_2back 0.8979 1.1523 19
Vavg_3back 1.3825 1.3182 19
Vavg_4back 1.6354 1.6525 19
72
Figure 20: Boxplot of Oxygenation Levels at Right DLPFC
Figure 21: Boxplot of Oxygenation Levels for Average of Left and Right
DLPFC
73
Repeated measures ANOVA with gender as a between-subjects factor and task type
(maze, IST-1, IST-2, 1-back, 2-back, 3-back and 4-back) as a within-subjects factor
was conducted to compare any cognitive task with other tasks to see any significant
difference exists. Repeated measures ANOVA was applied three times for left
DLPFC, right DLPFC and average oxygenation level of right and left DLPFC.
Results of test showed that there was an overall significant difference between
different tasks in oxygenation levels for left DLPFC (F =7.439, df =2.403, p <.05),
for right DLPFC (F =4.848, df =2.6723, p <.05) and for average oxygenation level
of right and left DLPFC (F =7.221, df =2.516, p <.05). Then, tasks were compared
with each other specifically.
First, maze task was compared with other tasks. Results (Table 26) show that, there
were significant differences in oxygenation levels at left DLPFC between maze and
IST-2 task (t(19)=-2.221, p<.05), maze and 3-back task (t(19)=-0.236, p<.05), maze
and 4-back task (t(19)=-0.202, p<.05). However, oxygenation levels at left DLPFC
do not differ significantly between maze and IST-1 task (t(19)=0.497, p>.05), maze
and 1-back (t(19)=1.636, p>.05), maze and 2-back (t(19)=-1.217, p>.05). Paired t-
test for oxygenation level at right DLPFC is shown in Table 27. The oxygenation
levels are significantly different between maze and IST-2 tasks (t(19)=-2.601,
p<.05), maze and 3-back tasks (t(19)=-2.356, p<.05) and maze and 4-back tasks
(t(19)=-2.853, p<.05). There were no significant differences between maze and
IST-1 tasks, maze and 1-back tasks and maze and 2-back tasks. The average of
oxygenation levels shows same results with left and right DLPFC regions as seen
from Table 28. While there were significant differences between maze and IST-2
tasks (t(19)=-2.601, p<.05), maze and 3-back tasks (t(19)=-2.356, p<.05), maze and
4-back tasks (t(19)=-2.853, p<.05); there were not any significant differences
between maze and IST-1 (t(19)= -0.335, p>.05), maze and 1-back (t(19)=0.997,
p>.05), maze and 2-back tasks (t(19)= -1.155, p>.05).
74
Table 26: Paired Samples Test to Compare Maze Task with others for Left
DLPFC
Table 27: Paired Samples Test to compare Maze Task with others for Right
DLPFC
Table 28: Paired Samples Test to compare Maze Task with Others for Average
Lower Upper
Pair 1 V1_maze -
V1_ist1.128 1.126 .258 -.415 .671 .497 19 .625
Pair 2 V1_maze -
V1_ist2-.581 1.140 .261 -1.130 -.031 -2.221 19 .039
Pair 3 V1_maze -
V1_1back.528 1.358 .312 -.126 1.183 1.696 19 .107
Pair 4 V1_maze -
V1_2back-.382 1.369 .314 -1.042 .278 -1.217 19 .239
Pair 5 V1_maze -
V1_3back-.955 1.492 .342 -1.674 -.236 -2.790 19 .012
Pair 6 V1_maze -
V1_4back-1.152 1.971 .452 -2.102 -.202 -2.548 19 .020
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviati
on
Std.
Error
Mean
95% Confidence
Interval of the
Difference
Lower Upper
Pair 1 V4_maze -
V4_ist1-.091 1.182 .271 -.661 .479 -.335 19 .742
Pair 2 V4_maze -
V4_ist2-.845 1.417 .325 -1.528 -.162 -2.601 19 .018
Pair 3 V4_maze -
V4_1back.276 1.206 .277 -.305 .857 .997 19 .332
Pair 4 V4_maze -
V4_2back-.361 1.363 .313 -1.018 .296 -1.155 19 .263
Pair 5 V4_maze -
V4_3back-.758 1.402 .322 -1.434 -.082 -2.356 19 .030
Pair 6 V4_maze -
V4_4back-1.067 1.630 .374 -1.853 -.281 -2.853 19 .011
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviati
on
Std.
Error
Mean
95% Confidence
Interval of the
Difference
75
of Left and Right DLPFC
Second, IST-1 and IST-2 were compared with other tasks to see whether any
significant difference exists. Paired t-test for oxygenation level at left DLPFC is
shown in Table 29. IST-1 task is significantly different from IST-2 task (t(19)=-
5.047, p<.001), 3-back task (t(19)=-3.810 , p<.05) and 4-back task (t(19)=-3.298,
p<.05) and IST-2 task is significantly different from maze (t(19)=2.628, p<.05),
IST-1 task (t(19)=-5.047, p<.001) and 1-back task (t(19)=3.713, p<.05); while there
were no significant differences between IST-1 and maze tasks (t(19)=0.05, p>.05),
IST-1 and 1-back tasks (t(19)=1.582, p>.05), IST-1 and 2-back tasks (t(19)=-2.038,
p>.05), IST-2 and 2-back tasks (t(19)=0.708, p>.05), IST-2 and 3-back tasks
(t(19)=-1.307, p>.05), IST-2 and 4-back tasks (t(19)=-1.434, p>.05).
Paired t-tests for right DLPFC is presented in Table 30 and results of right DLPFC is
Lower Upper
Pair 1 Vavg_maze -
Vavg_ist1.024 1.018 .228 -.453 .500 .105 19 .918
Pair 2 Vavg_maze -
Vavg_ist2-.779 1.238 .277 -1.359 -.200 -2.814 19 .011
Pair 3 Vavg_maze -
Vavg_1back.391 1.175 .270 -.176 .957 1.449 19 .165
Pair 4 Vavg_maze -
Vavg_2back-.372 1.259 .289 -.978 .235 -1.288 19 .214
Pair 5 Vavg_maze -
Vavg_3back-.856 1.343 .308 -1.504 -.209 -2.779 19 .012
Pair 6 Vavg_maze -
Vavg_4back-1.109 1.698 .390 -1.928 -.291 -2.848 19 .011
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviati
on
Std.
Error
Mean
95% Confidence
Interval of the
Difference
76
same as left one. IST-1 task is significantly different from IST-2 task (t(19)=-4.030,
p<.05), 3-back task (t(19)=-2.558, p<.05) and 4-back task (t(19)=-2.948, p<.05) and
IST-2 task is significantly different from maze task (t(19)=2.601, p<.05), IST-1 task
(t(19)=4.030, p<.05) and 1-back task (t(19)=3.763, p<.05); while there are no
significantly differences between IST-1 and maze tasks (t(19)=0.05, p>.05), IST-1
and 1-back tasks (t(19)=1.754, p>.05), IST-1 and 2-back tasks (t(19)=-0.983,
p>.05), IST-2 and 2-back tasks (t(19)=-0.983, p>.05), IST-2 and 3-back tasks
(t(19)=0.235, p>.05), IST-2 and 4-back tasks (t(19)=-0.498, p>.05).
Since paired t-test for left and right DLPFC regions show same results their average
will show the same. Despite knowing that, paired t-test was also conducted for
average oxygenation levels as seen from Table 31.
77
Table 29: Paired Samples Test to Compare IST with others for Left DLPFC
Lower Upper
Pair 1 V1_ist1 -
V1_maze.014 1.210 .278 -.569 .597 .050 19 .961
Pair 2 V1_ist1 -
V1_ist2-.709 .612 .140 -1.004 -.414 -5.047 19 .000
Pair 3 V1_ist1 -
V1_1back.400 1.102 .253 -.131 .931 1.582 19 .131
Pair 4 V1_ist1 -
V1_2back-.511 1.093 .251 -1.037 .016 -2.038 19 .057
Pair 5 V1_ist1 -
V1_3back-1.083 1.240 .284 -1.681 -.486 -3.810 19 .001
Pair 6 V1_ist1 -
V1_4back-1.280 1.692 .388 -2.096 -.465 -3.298 19 .004
Pair 7 V1_ist2 -
V1_maze.723 1.199 .275 .145 1.301 2.628 19 .017
Pair 8 V1_ist2 -
V1_ist1.709 .612 .140 .414 1.004 5.047 19 .000
Pair 9 V1_ist2 -
V1_1back1.109 1.302 .299 .481 1.736 3.713 19 .002
Pair 10 V1_ist2 -
V1_2back.198 1.220 .280 -.390 .786 .708 19 .488
Pair 11 V1_ist2 -
V1_3back-.374 1.249 .286 -.976 .227 -1.307 19 .208
Pair 12 V1_ist2 -
V1_4back-.571 1.737 .398 -1.409 .266 -1.434 19 .169
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviatio
n
Std.
Error
Mean
95% Confidence
Interval of the
Difference
78
Table 30: Paired Samples Test to Compare IST with others for Right DLPFC
Lower Upper
Pair 1 V4_ist1 -
V4_maze.014 1.210 .278 -.569 .597 .050 18 .961
Pair 2 V4_ist1 -
V4_ist2-.755 .816 .187 -1.148 -.361 -4.030 18 .001
Pair 3 V4_ist1 -
V4_1back.367 .911 .209 -.073 .806 1.754 18 .096
Pair 4 V4_ist1 -
V4_2back-.270 1.199 .275 -.848 .307 -.983 18 .338
Pair 5 V4_ist1 -
V4_3back-.667 1.136 .261 -1.215 -.119 -2.558 18 .020
Pair 6 V4_ist1 -
V4_4back-.976 1.443 .331 -1.672 -.280 -2.948 18 .009
Pair 7 V4_ist2 -
V4_maze.845 1.417 .325 .162 1.528 2.601 18 .018
Pair 8 V4_ist2 -
V4_ist1.755 .816 .187 .361 1.148 4.030 18 .001
Pair 9 V4_ist2 -
V4_1back1.121 1.299 .298 .495 1.747 3.763 18 .001
Pair 10 V4_ist1 -
V4_2back-.270 1.199 .275 -.848 .307 -.983 18 .338
Pair 11 V4_ist2 -
V4_3back.088 1.626 .373 -.696 .871 .235 18 .817
Pair 12 V4_ist2 -
V4_4back-.221 1.939 .445 -1.156 .713 -.498 18 .625
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviatio
n
Std.
Error
Mean
95% Confidence
Interval of the
Difference
79
Table 31: Paired Samples Test to Compare IST with others for Average of Left
and Right DLPFC
Lastly, n-back tasks were compared with other tasks. Paired t-test for oxygenation
level at left DLPFC is shown in Table 32. 1-back task is significantly different from
IST-2 task (t(19)=-3.713, p<.05), 2-back task (t(19)=-5.909, p<.001), 3-back task
Lower Upper
Pair 1 Vavg_ist1 -
Vavg_maze-.024 1.018 .228 -.500 .453 -.105 19 .918
Pair 2 Vavg_ist1 -
Vavg_ist2-.803 .621 .139 -1.094 -.512 -5.780 19 .000
Pair 3 Vavg_ist1 -
Vavg_1back.354 .947 .217 -.102 .811 1.631 18 .120
Pair 4 Vavg_ist1 -
Vavg_2back-.408 1.026 .235 -.903 .086 -1.734 18 .100
Pair 5 Vavg_ist1 -
Vavg_3back-.893 1.118 .257 -1.432 -.354 -3.480 18 .003
Pair 6 Vavg_ist1 -
Vavg_4back-1.146 1.469 .337 -1.854 -.438 -3.399 18 .003
Pair 7 Vavg_ist2 -
Vavg_maze.779 1.238 .277 .200 1.359 2.814 19 .011
Pair 8 Vavg_ist2 -
Vavg_ist1.803 .621 .139 .512 1.094 5.780 19 .000
Pair 9 Vavg_ist2 -
Vavg_1back1.138 1.283 .294 .519 1.756 3.865 18 .001
Pair 10 Vavg_ist2 -
Vavg_2back.375 1.312 .301 -.257 1.008 1.247 18 .228
Pair 11 Vavg_ist2 -
Vavg_3back-.109 1.356 .311 -.763 .544 -.351 18 .730
Pair 12 Vavg_ist2 -
Vavg_4back-.362 1.713 .393 -1.188 .464 -.922 18 .369
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviatio
n
Std.
Error
Mean
95% Confidence
Interval of the
Difference
80
(t(19)=-7.110, p<.001) and 4-back task (t(19)=-6.627, p<.001); 2-back task is
significantly different from 1-back (t(19)=-5.909, p<.001), 3-back (t(19)=5.062,
p<.001) and 4-back (t(19)=-3.297, p<.05) tasks; 3-back is significantly different
from IST-1 task (t(19)=3.810, p<.05), maze (t(19)=2.790, p<.05), 1-back (t(19)=-
7.110, p<.001) and 2-back tasks (t(19)=5.062, p<.001); 4-back is significantly
different from IST-1 (t(19)=3.298, p<.05), maze (t(19)=2.548, p<.05), 1-back
(t(19)=-6.627, p<.001) and 2-back (t(19)=-3.297, p<.05) tasks.
Paired t-test for oxygenation level of right DLPFC was conducted and is presented
in Table 33. 1-back task is significantly different from IST-2 (t(19)=-3.763, p<.05),
2-back (t(19)=-3.587, p<.05), 3-back (t(19)=,-4.706 p<.001) and 4-back (t(19)=-
4.712, p<.001) tasks; 2-back task is significantly different from 1-back (t(19)=-
2.457, p<.05), 3-back (t(19)=-4.706, p<.05) and 4-back (t(19)=-2.894, p<.05) tasks;
3-back is significantly different from IST-1 task (t(19)=2.558, p<.05), maze
(t(19)=2.356, p<.05), 1-back (t(19)=-4.706, p<.001), 2-back (t(19)=-2.457, p<.05)
and 4-back (t(19)=-2.178, p<.05) tasks; 4-back is significantly different from IST-1
(t(19)=2.948, p<.05), maze (t(19)=2.853, p<.05), 1-back (t(19)=-4.712, p<.001), 2-
back (t(19)=-2.894, p<.05) and 3-back (t(19)=-2.178, p<.05) tasks.
Even though paired t-tests for right and left DLPFC are similar, they are not the
same. Therefore, another paired t-test was conducted for their average oxygenation
levels (Table 34). Results show that 1-back task is significantly different from IST-2
(t(19)=,-1.631 p<.05), 2-back (t(19)=-5.167, p<.05), 3-back (t(19)=-6.532, p<.05)
and 4-back (t(19)=-6.367, p<.05) tasks; 2-back task is significantly different from 1-
back (t(19)=-5.167, p<.05), 3-back (t(19)=-4.046, p<.05) and 4-back (t(19)=3.465,
p<.05) tasks; 3-back is significantly different from IST-1 task (t(19)=3.480, p<.05),
maze (t(19)=2.779, p<.05), 1-back (t(19)=6.532, p<.05) and 2-back (t(19)=4.046,
p<.05) tasks; 4-back is significantly different from IST-1 (t(19)=3.399, p<.05), maze
(t(19)=2.848, p<.05), 1-back (t(19)=6.367, p<.05) and 2-back (t(19)=3.465, p<.05)
tasks .
81
Table 32: Paired Samples Test for compare N-back Tasks with Others for Left
DLPFC
Lower Upper
Pair 1 V1_1back -
V1_ist1-.400 1.102 .253 -.931 .131 -1.582 19 .131
Pair 2 V1_1back -
V1_ist2-1.109 1.302 .299 -1.736 -.481 -3.713 19 .002
Pair 3 V1_1back -
V1_maze-.528 1.358 .312 -1.183 .126 -1.696 19 .107
Pair 4 V1_1back -
V1_2back-.911 .672 .154 -1.235 -.587 -5.909 19 .000
Pair 5 V1_1back -
V1_3back-1.483 .909 .209 -1.922 -1.045 -7.110 19 .000
Pair 6 V1_1back -
V1_4back-1.680 1.105 .254 -2.213 -1.148 -6.627 19 .000
Pair 7 V1_2back -
V1_ist1.511 1.093 .251 -.016 1.037 2.038 19 .057
Pair 8 V1_2back -
V1_ist2-.198 1.220 .280 -.786 .390 -.708 19 .488
Pair 9 V1_2back -
V1_maze.382 1.369 .314 -.278 1.042 1.217 19 .239
Pair 10 V1_2back -
V1_1back.911 .672 .154 .587 1.235 5.909 19 .000
Pair 11 V1_2back -
V1_3back-.573 .493 .113 -.810 -.335 -5.062 19 .000
Pair 12 V1_2back -
V1_4back-.769 1.017 .233 -1.260 -.279 -3.297 19 .004
Pair 13 V1_3back -
V1_ist11.083 1.240 .284 .486 1.681 3.810 19 .001
Pair 14 V1_3back -
V1_ist2.374 1.249 .286 -.227 .976 1.307 19 .208
Pair 15 V1_3back -
V1_maze.955 1.492 .342 .236 1.674 2.790 19 .012
Pair 16 V1_3back -
V4_1back1.373 1.101 .252 .842 1.903 5.438 19 .000
Pair 17 V1_3back -
V1_2back.573 .493 .113 .335 .810 5.062 19 .000
Pair 18 V1_3back -
V1_4back-.197 .837 .192 -.600 .206 -1.026 19 .319
Pair 19 V1_4back -
V1_ist11.280 1.692 .388 .465 2.096 3.298 19 .004
Pair 20 V1_4back -
V1_ist2.571 1.737 .398 -.266 1.409 1.434 19 .169
Pair 21 V1_4back -
V1_maze1.152 1.971 .452 .202 2.102 2.548 19 .020
Pair 22 V1_4back -
V1_1back1.680 1.105 .254 1.148 2.213 6.627 19 .000
Pair 23 V1_4back -
V1_2back.769 1.017 .233 .279 1.260 3.297 19 .004
Pair 24 V1_4back -
V1_3back.197 .837 .192 -.206 .600 1.026 19 .319
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviation
Std. Error
Mean
95% Confidence
Interval of the
Difference
82
Table 33: Paired Samples Test for compare N-back Tasks with Others for
Right DLPFC
Lower Upper
Pair 1 V4_1back -
V4_ist1-.367 .911 .209 -.806 .073 -1.754 19 .096
Pair 2 V4_1back -
V4_ist2-1.121 1.299 .298 -1.747 -.495 -3.763 19 .001
Pair 3 V4_1back -
V4_maze-.276 1.206 .277 -.857 .305 -.997 19 .332
Pair 4 V4_1back -
V4_2back-.637 .774 .178 -1.010 -.264 -3.587 19 .002
Pair 5 V4_1back -
V4_3back-1.034 .957 .220 -1.495 -.572 -4.706 19 .000
Pair 6 V4_1back -
V4_4back-1.343 1.242 .285 -1.941 -.744 -4.712 19 .000
Pair 7 V4_2back -
V4_ist1.270 1.199 .275 -.307 .848 .983 19 .338
Pair 8 V4_2back -
V4_ist2-.484 1.589 .365 -1.250 .282 -1.328 19 .201
Pair 9 V4_2back -
V4_maze.361 1.363 .313 -.296 1.018 1.155 19 .263
Pair 10 V4_2back -
V4_1back.637 .774 .178 .264 1.010 3.587 19 .002
Pair 11 V4_2back -
V4_3back-.397 .703 .161 -.736 -.057 -2.457 19 .024
Pair 12 V4_2back -
V4_4back-.706 1.063 .244 -1.218 -.193 -2.894 19 .010
Pair 13 V4_3back -
V4_ist1.667 1.136 .261 .119 1.215 2.558 19 .020
Pair 14 V4_3back -
V4_ist2-.088 1.626 .373 -.871 .696 -.235 19 .817
Pair 15 V4_3back -
V4_maze.758 1.402 .322 .082 1.434 2.356 19 .030
Pair 16 V4_3back -
V4_1back1.034 .957 .220 .572 1.495 4.706 19 .000
Pair 17 V4_3back -
V4_2back.397 .703 .161 .057 .736 2.457 19 .024
Pair 18 V4_3back -
V4_4back-.309 .619 .142 -.607 -.011 -2.178 19 .043
Pair 19 V4_4back -
V4_ist1.976 1.443 .331 .280 1.672 2.948 19 .009
Pair 20 V4_4back -
V4_ist2.221 1.939 .445 -.713 1.156 .498 19 .625
Pair 21 V4_4back -
V4_maze1.067 1.630 .374 .281 1.853 2.853 19 .011
Pair 22 V4_4back -
V4_1back1.343 1.242 .285 .744 1.941 4.712 19 .000
Pair 23 V4_4back -
V4_2back.706 1.063 .244 .193 1.218 2.894 19 .010
Pair 24 V4_4back -
V4_3back.309 .619 .142 .011 .607 2.178 19 .043
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviation
Std. Error
Mean
95% Confidence
Interval of the
Difference
83
Table 34: Paired Samples Test for compare N-back Tasks with Others for
Average of Left and Right DLPFC
Lower Upper
Pair 1 Vavg_1back -
Vavg_maze-.391 1.175 .270 -.957 .176 -1.449 19 .165
Pair 2 Vavg_1back -
Vavg_ist1-.354 .947 .217 -.811 .102 -1.631 19 .120
Pair 3 Vavg_1back -
Vavg_ist2-1.138 1.283 .294 -1.756 -.519 -3.865 19 .001
Pair 4 Vavg_1back -
Vavg_2back-.763 .643 .148 -1.073 -.452 -5.167 19 .000
Pair 5 Vavg_1back -
Vavg_3back-1.247 .832 .191 -1.648 -.846 -6.532 19 .000
Pair 6 Vavg_1back -
Vavg_4back-1.500 1.027 .236 -1.995 -1.005 -6.367 19 .000
Pair 7 Vavg_2back -
Vavg_maze.372 1.259 .289 -.235 .978 1.288 19 .214
Pair 8 Vavg_2back -
Vavg_ist1.408 1.026 .235 -.086 .903 1.734 19 .100
Pair 9 Vavg_2back -
Vavg_ist2-.375 1.312 .301 -1.008 .257 -1.247 19 .228
Pair 10 Vavg_2back -
Vavg_1back.763 .643 .148 .452 1.073 5.167 19 .000
Pair 11 Vavg_2back -
Vavg_3back-.485 .522 .120 -.736 -.233 -4.046 19 .001
Pair 12 Vavg_2back -
Vavg_4back-.738 .928 .213 -1.185 -.290 -3.465 19 .003
Pair 13 Vavg_3back -
Vavg_maze.856 1.343 .308 .209 1.504 2.779 19 .012
Pair 14 Vavg_3back -
Vavg_ist1.893 1.118 .257 .354 1.432 3.480 19 .003
Pair 15 Vavg_3back -
Vavg_ist2.109 1.356 .311 -.544 .763 .351 19 .730
Pair 16 Vavg_3back -
Vavg_1back1.247 .832 .191 .846 1.648 6.532 19 .000
Pair 17 Vavg_3back -
Vavg_2back.485 .522 .120 .233 .736 4.046 19 .001
Pair 18 Vavg_3back -
Vavg_4back-.253 .648 .149 -.565 .060 -1.700 19 .106
Pair 19 Vavg_4back -
Vavg_maze1.109 1.698 .390 .291 1.928 2.848 19 .011
Pair 20 Vavg_4back -
Vavg_ist11.146 1.469 .337 .438 1.854 3.399 19 .003
Pair 21 Vavg_4back -
Vavg_ist2.362 1.713 .393 -.464 1.188 .922 19 .369
Pair 22 Vavg_4back -
Vavg_1back1.500 1.027 .236 1.005 1.995 6.367 19 .000
Pair 23 Vavg_4back -
Vavg_2back.738 .928 .213 .290 1.185 3.465 19 .003
Pair 24 Vavg_4back -
Vavg_3back.253 .648 .149 -.060 .565 1.700 19 .106
Paired Differences
t df
Sig. (2-
tailed)Mean
Std.
Deviation
Std. Error
Mean
95% Confidence
Interval of the
Difference
84
To sum up, from low oxygenation level to highest, tasks can be ranked as follows:
1-back, IST-1, maze, 2-back, IST-2, 3-back, 4-back in voxel-1; 1-back, maze, IST-1,
2-back, IST-2, 3-back, 4-back in voxel-4; and lastly 1-back, IST-1, maze, 2-back,
IST-2, 3-back, 4-back at average oxygenation levels of left and right DLPFC
regions. These rankings and significant differences give an idea about task difficulty
of different kind of cognitive tasks. For example, when IST-1 is compared with n-
back tasks difficulty of the task is between 1-back and 2-back according to rankings
and paired t-test show actually there was no significant differences in their blood
oxygenation levels. Therefore, we can conclude that the difficulty of the task is
similar with 1-back and 2-back.
Blood oxygenation level provide assessing mental workload of participants, so
comparing these oxygenation levels during cognitive tasks gives information about
difficulties of tasks.
4.6. The Relationship of fNIR Measures
fNIR data was collected from all 37 participants, but physical problems (different
shapes of foreheads) raise difficulties in collecting fNIR data from all participants
from all voxels. Location of voxels is shown in Figure 22. There are 2 detectors on
each fNIR sensor and that means there are 4 channels provide hemoglobin and oxy-
hemoglobin level data. After filtering fNIR data by FIR and SMAR, 19 participants’
data of voxel-1 and voxel-4 were available to be used.
85
Figure 22: Location of Voxels at Prefrontal Cortex
Since voxel-1 is gathering information about situation of left DLPFC and voxel-4 is
gathering information about situation of right DLPFC, first research question of
fNIR data about whether there is a difference between changes in oxygenation levels
of left DLPFC and right DLPFC during a cognitive task. Paired t-test were applied
for 19 participants (8 females and 11 males) for each cognitive tasks and results
show that there was no significant differences between changes in oxygenation
levels of left DLPFC and right DLPFC during maze task, IST-1 and IST-2 tasks, and
all n-back tasks since p>.05. Means and standard deviations of oxygenation levels
can be seen from Table 35.The result of paired t-test is shown in Table 36.
86
Table 35: Descriptive Statistics for Oxygenation Levels at Left and Right
DLPFC
Table 36: Paired t-Test for Oxygenation Levels at Left and Right DLPFC
Mean N
Std.
Deviation
Std. Error
Mean
V1_maze (Left) .60 19 1.034 0.237
V4_maze (Right) .45 19 0.879 0.202
V1_ist1 (Left) .47 19 0.888 0.204
V4_ist1 (Right) .55 19 1.005 0.231
V1_ist2 (Left) 1.18 19 0.929 0.213
V4_ist2 (Right) 1.30 19 1.036 0.238
V1_1back (Left) .07 19 1.096 0.251
V4_1back (Right) .18 19 0.896 0.206
V1_2back (Left) .98 19 1.217 0.279
V4_2back (Right) .82 19 1.341 0.308
V1_3back (Left) 1.55 19 1.350 0.310
V4_3back (Right) 1.21 19 1.440 0.330
V1_4back (Left) 1.75 19 1.815 0.416
V4_4back (Right) 1.52 19 1.675 0.384
Pair 6
Pair 7
Pair 1
Pair 2
Pair 3
Pair 4
Pair 5
Lower Upper
Pair 1 V1_maze - V4_maze .142 1.088 .250 -.382 .666 .570 19 .576
Pair 2 V1_ist1 - V4_ist1 -.077 .906 .208 -.514 .360 -.370 19 .715
Pair 3 V1_ist2 - V4_ist2 -.123 .778 .178 -.498 .252 -.687 19 .501
Pair 4 V1_1back - V4_1back -.110 .856 .196 -.523 .302 -.562 19 .581
Pair 5 V1_2back - V4_2back .163 1.117 .256 -.375 .702 .637 19 .532
Pair 6 V1_3back - V4_3back .339 .916 .210 -.102 .781 1.614 19 .124
Pair 7 V1_4back - V4_4back .227 1.130 .259 -.318 .772 .876 19 .392
Paired Differences
t df
Sig. (2-
tailed)Mean Std. Deviation
Std.
Error
Mean
95%
Confidence
Interval of the
Difference
87
4.7. Correlation between fNIR and Performance Measures
Another research question is about the relationship between fNIR and performance
measures. Therefore, correlations between measures were investigated for each task.
Pearson correlation was applied. Maze path length and maze time are two
performance measures for maze task and the oxygenation level at right DLPFC and
the oxygenation level at left DLPFC are two fNIR measures for maze. Four different
measures were compared by Pearson correlation and results are shown in Table 37.
Since p<.05, there is a significant linear relationship between maze path length and
maze time. The Pearson correlation is 0.816. This means there is a strong positive
relationship between maze path length and time. Since other p values are greater
than .05, it cannot be said that there is any correlation between them.
Table 37: Correlations between Maze Measures
MazePath MazeTime V1_maze V4_maze
Pearson 1 0.816** 0.078 -0.113
Sig. (2-tailed) 0 0.752 0.646
N 19 19 19 19
Pearson 0.816** 1 -0.064 0.03
Sig. (2-tailed) 0 0.796 0.903
N 19 19 19 19
Pearson 0.078 -0.064 1 0.363
Sig. (2-tailed) 0.752 0.796 0.127
N 19 19 19 19
Pearson -0.113 0.03 0.363 1
Sig. (2-tailed) 0.646 0.903 0.127
N 19 19 19 19
MazePath
MazeTime
V1_maze
V4_maze
88
Performance, trial number and completion time are used as performance measures
and the oxygenation level at right DLPFC and the oxygenation level at left DLPFC
are two fNIR measures for ISTs. The result of Pearson correlation for IST-1 is
presented in Table 38. There are significant relationships between performance and
trial number, performance and time, trial number and time, the oxygenation level at
right DLPFC and the oxygenation level at left DLPFC since p<.05. Pearson
correlation between IST-1 performance and IST-1 Trial number is 0.717, so there is
a strong positive correlation between them. Pearson correlation between IST-1
performance and IST-1 time is 0.508, so there is a strong positive correlation
between them. Pearson correlation between IST-1 trial number and IST-1 time is
0.768, so there is a strong positive correlation between them. Pearson correlation
between the oxygenation level in voxel-1 and voxel-4 is 0.548, so there is a strong
positive correlation between them (Table 38).
Table 38: Correlations between IST-1 Measures
IST1-Per IST1-TrialNo IST1-Time V1_ist1 V4_ist1
Pearson
Correlation
1 0.717**
0.508* -0.184 -0.200
Sig. (2-tailed) 0.001 0.026 0.452 0.411
N 19 19 19 19 19
Pearson
Correlation0.717
** 1 0.768** 0.004 -0.314
Sig. (2-tailed) 0.001 0.000 0.986 0.190
N 19 19 19 19 19
Pearson
Correlation
0.508*
0.768** 1 -0.183 -0.252
Sig. (2-tailed) 0.026 0.000 0.453 0.298
N 19 19 19 19 19
Pearson
Correlation
-0.184 0.004 -0.183 1 0.548*
Sig. (2-tailed) 0.452 0.986 0.453 0.015
N 19 19 19 19 19
Pearson
Correlation
-0.200 -0.314 -0.252 0.548* 1
Sig. (2-tailed) 0.411 0.190 0.298 0.015
N 19 19 19 19 19
V4_ist1
IST1-Per
IST1-TrialNo
IST1-Time
V1_ist1
89
The result of Pearson correlation for IST-2 to investigate relationship of 5 measures
is presented in Table 39. There are significant relationships between performance
and trial number, trial number and time, the oxygenation level at right DLPFC and
the oxygenation level at left DLPFC since p<.05. Pearson correlation between IST-2
performance and IST-2 Trial number is -0.499, so there is a negative moderate
correlation between them. Pearson correlation between IST-2 trial number and IST-
2 time is 0.809, so there is a strong positive correlation between them. Pearson
correlation between the oxygenation level in voxel-1 and voxel-4 is 0.692, so there
is a strong positive correlation between them (Table 39).
Table 39: Correlations between IST-2 Measures
IST2-Per IST2-TrialNo IST2-Time V1_ist2 V4_ist2
Pearson
Correlation
1 -0.499 -0.304 -0.386 -0.280
Sig. (2-tailed) 0.030 0.206 0.103 0.245
N 19 19 19 19 19
Pearson
Correlation
-0.499 1 0.809** -0.029 -0.094
Sig. (2-tailed) 0.030 0.000 0.908 0.703
N 19 19 19 19 19
Pearson
Correlation
-0.304 0.809** 1 0.042 0.008
Sig. (2-tailed) 0.206 0.000 0.865 0.975
N 19 19 19 19 19
Pearson
Correlation
-0.386 -0.029 0.042 1 0.692**
Sig. (2-tailed) 0.103 0.908 0.865 0.001
N 19 19 19 19 19
Pearson
Correlation
-0.280 -0.094 0.008 0.692** 1
Sig. (2-tailed) 0.245 0.703 0.975 0.001
N 19 19 19 19 19
V1_ist2
V4_ist2
IST2-Per
IST2-TrialNo
IST2-Time
90
Performance is one and only performance measure and the oxygenation level at right
DLPFC and the oxygenation level at left DLPFC are two fNIR measures for n-back
tasks. Four different Pearson correlations were conducted to see relationship for
each task. As seen from Table 40, there is a significant relationship between the
oxygenation level at right DLPFC and the oxygenation level at left DLPFC since
p<.05 but performance does not have any relationship with fNIR measure at .05
level. Pearson correlation between the oxygenation level in voxel-1 and voxel-4 is
0.657, so there is a strong positive correlation between them.
Pearson correlation for 2-back task can be seen from Table 41 and results show that
performance as performance measure does not have a significant relationship with
fNIR measures while there is a significant relationship the oxygenation level at right
DLPFC and the oxygenation level at left DLPFC since p<.05. Pearson correlation
between the oxygenation level in voxel-1 and voxel-4 is 0.647, so there is a strong
positive correlation between them.
Table 40: Correlations between 1-back Measures
back1 V1_1back V4_1back
Pearson 1 0.084 0.022
Sig. (2-tailed) 0.732 0.928
N 19 19 19
Pearson 0.084 1 0.647**
Sig. (2-tailed) 0.732 0.003
N 19 19 19
Pearson 0.022 0.647** 1
Sig. (2-tailed) 0.928 0.003
N 19 19 19
V4_1back
back1
V1_1back
91
Table 41: Correlations between 2-back Measures
Pearson correlation for 3-back task can be seen from Table 42 and results show that
performance as performance measure does not have a significant relationship with
fNIR measures while there is a significant relationship the oxygenation level at right
DLPFC and the oxygenation level at left DLPFC since p<.05. Pearson correlation
between the oxygenation level in voxel-1 and voxel-4 is 0.786, so there is a strong
positive correlation between them.
back2 V1_2back V4_2back
Pearson 1 0.048 0.128
Sig. (2-tailed) 0.846 0.603
N 19 19 19
Pearson 0.048 1 0.623**
Sig. (2-tailed) 0.846 0.004
N 19 19 19
Pearson 0.128 0.623** 1
Sig. (2-tailed) 0.603 0.004
N 19 19 19
back2
V1_2back
V4_2back
92
Table 42: Correlations between 3-back Measures
Pearson correlation for 4-back task can be seen from Table 43 and results show that
performance as a performance measure does not have a significant relationship with
fNIR measures while there is a significant relationship between the oxygenation
level at right DLPFC and the oxygenation level at left DLPFC since p<.05. Pearson
correlation between the oxygenation level in voxel-1 and voxel-4 is 0.534, so there
is a strong positive correlation between them.
back3 V1_3back V4_3back
Pearson 1 -0.051 0.137
Sig. (2-tailed) 0.834 0.577
N 19 19 19
Pearson -0.051 1 0.786**
Sig. (2-tailed) 0.834 0.000
N 19 19 19
Pearson 0.137 0.786** 1
Sig. (2-tailed) 0.577 0.000
N 19 19 19
back3
V1_3back
V4_3back
93
Table 43: Correlations between 4-back Measures
One of the research questions of this study is whether there is a relationship between
performance performance and fNIR results. Correlations show us there is no strong
relationship between behavior and brain activations. In the literature, there are not
any previous studies or any evidence for their relationship because it is not expected.
Showing high performance in a task does not mean she/he shows high brain
activation or vice versa. Thus, this result is not surprising.
4.8. Personality Traits
A Turkish culture based basic personality traits inventory (Gencoz and Oncul, 2012)
was used in our study to observe personality differences. The test investigated a
person under 6 different dimensions which were extraversion, conscientiousness,
agreeableness, neuroticism, openness to experience and negative valence. The
results of this questionnaire gave a value between 1 and 5 for each trait. For
example, a person who had a point close to 1 for extraversion dimension, show less
back4 V1_4back V4_4back
Pearson 1 0.397 0.152
Sig. (2-tailed) 0.093 0.534
N 19 19 19
Pearson 0.397 1 0.793**
Sig. (2-tailed) 0.093 0.000
N 19 19 19
Pearson 0.152 0.793** 1
Sig. (2-tailed) 0.534 0.000
N 19 19 19
back4
V1_4back
V4_4back
94
extraverted features while a point close to 5 means more extraverted characteristics.
In our study, we used the experimental data of Gencoz and Oncul (2012)’s study to
have an idea about average value on personality trait of Turkish people. 454
participants were joined their study and descriptive statistics are shown in Table 44.
The mean for each trait was taken as a base point and personality traits data in our
study was classified lower or higher than that value. The value for extraversion
personality trait is 3.47. If a participant had higher value than 3.47 for extraversion
dimension, “higher” was written down for that trait; if she/he did not, “lower” was
recorded. This higher/lower classification is given in Table 45. While agreeableness,
neuroticism and openness to experience traits are divided into two groups
homogeneously; more people with less negative valence and more conscientiousness
are participated the experiment. Finding participants from wider area might have
prevented the deviation.
Before starting any analyses, internal consistency of Basic Personality Traits
Inventory was conducted by using Cronbach’s alpha to ensure the reliability of
participants’ self-questionaries’ personality traits. Cronbach's alphas for each
dimension were high: 0.86 for extraversion, 0.89 for conscientiousness, 0.86 for
agreeableness, 0.80 for neuroticism, 0.79 for openness to experience and 0.72 for
negative valence. Results show that Cranbach’s alphas were higher than 0.7 for each
personality factor. Therefore, it can be concluded that they are reliable data to work
on.
Descriptive statistics for personality traits in our study were also computed to see
how participants of our study differ from ones at Gencoz and Oncul (2012). It is
seen that mean values of traits are very close to each other. Descriptive statistics for
personality trait in our experiment are given in Appendix P. Detailed results for each
dimension and for each participant, and higher and lower version of results can also
be seen in Appendix R and Appendix S, respectively.
95
Table 44: Descriptive Statistics for Personality Traits in Turkey (Gencoz and
Oncul, 2012)
Table 45: Higher and Lower Classification for Personality in the Experiment
4.9. Personality Differences in Performance Measures
Pearson correlation coefficient was calculated to find out whether there is a
correlation between personality traits (extraversion, conscientiousness,
agreeableness, neuroticism, openness to experience and negative valence) and
N Minimum Maximum Mean Std. Deviation
Extraversion 454 1.25 5.00 3.47 0.82
Conscientiousness 454 1.13 5.00 3.45 0.75
Agreeableness 454 2.25 5.00 4.13 0.51
Neuroticism 454 1.00 4.78 2.78 0.74
Openness 454 1.33 5.00 3.66 0.68
Negative valence 454 1.00 4.17 1.69 0.55
Personality Trait N N
lower 16
higher 21
lower 24
higher 13
lower 20
higher 17
lower 19
higher 18
lower 18
higher 19
lower 27
higher 10
Neuroticism37
Openness37
Negative valence37
Extraversion37
Conscientiousness37
Agreeableness37
96
performance measures (maze path length, maze time, IST-1 completion time, IST-1
performance, IST-1 trial number, IST-2 completion time, IST-2 performance, IST-2
trial number and n-back performances).
First, the correlation was conducted for maze performance measures. Results show
that we are not confident that there is a correlation between any personality trait and
any maze performance measures. Test summary and group statistics of each
personality trait for each maze performance measure can be found in Table 46 and
Table 47, respectively.
When the relationship between personality traits and maze task is searched in the
literature, results show that no correlation between maze scores and two personality
traits (emotional stability and sociability) was found in Davies (1965). Emotional
stability stands for negative meaning of neuroticism in the personality traits that was
used in our experiment. Sociability has closer meaning to extraversion and openness
to experience traits. Therefore, literature has similar finding with our study about
their relationship.
97
Table 46: Test Summary of All Personality Traits for Maze Performance
Measures
Maze Path Maze Time
Pearson Correlation .055 .026
Sig. (2-tailed) .745 .879
N 37 37
Pearson Correlation -.173 -.066
Sig. (2-tailed) .307 .700
N 37 37
Pearson Correlation .020 .134
Sig. (2-tailed) .908 .428
N 37 37
Pearson Correlation .038 .089
Sig. (2-tailed) .822 .602
N 37 37
Pearson Correlation -.245 -.305
Sig. (2-tailed) .143 .067
N 37 37
Pearson Correlation -.006 -.205
Sig. (2-tailed) .973 .225
N 37 37
PF6 (Negative Valence)
PF1 (Extraversion)
PF2 (Conscientiousness)
PF3 (Agreableness)
PF4 (Neuroticism)
PF5 (Openness to
Experience)
98
Table 47: Group Statistics of All Personality Traits for Maze Performance
Measures
Personality Trait N Mean Std. DeviationStd. Error
Mean
lower 16 189 129.178 32.295
higher 21 198.1 122.395 26.709
lower 16 218.02 127.239 31.81
higher 21 239.52 99.988 21.819
lower 24 214.25 131.07 26.755
higher 13 157.08 103.271 28.642
lower 24 243.05 124.213 25.355
higher 13 206.54 82.104 22.772
lower 20 181.33 116.978 26.157
higher 17 209.27 133.112 32.284
lower 20 207.1 83.939 18.769
higher 17 257.43 134.595 32.644
lower 19 175.03 129.227 29.647
higher 18 214.36 117.767 27.758
lower 19 211.89 113.237 25.978
higher 18 249.57 109.294 25.761
lower 18 220.7 126.365 29.785
higher 19 169.03 118.915 27.281
lower 18 249.79 112.916 26.615
higher 19 211.68 109.753 25.179
lower 27 205.45 130.71 25.155
higher 10 163.7 102.058 32.274
lower 27 249.49 116.049 22.334
higher 10 178.2 81.483 25.767
PF1 (Extraversion)
PF2 (Conscientiousness)
PF3 (Agreeableness)
PF4 (Neuroticism)
PF5 (Openness to
Experience)
PF6 (Negative Valence)
MazePath
MazeTime
MazePath
MazeTime
MazePath
MazeTime
MazePath
MazeTime
Measurement
MazePath
MazeTime
MazePath
MazeTime
99
Pearson correlation was run for all IST performance variables (IST-1 completion
time, IST-1 performance, IST-1 trial number, IST-2 completion time, IST-2
performance, IST-2 trial number) to see whether there is a correlation between any
IST performance measure and any personality trait. Since r(37)= -.398, p≤.05; we
may conclude that a negative correlation between conscientiousness and IST-1
performance exists. There is also a negative correlation between agreeableness and
IST-1 trial number, where r(37)= -.341, p≤.05.
For other IST performance measures and personality traits, we cannot say that there
is a correlation between them. Test summary and group statistics of each personality
trait for each IST performance measure can be found in Table 48 and Table 49,
respectively.
Trial number in IST show risk-taking measure during a decision making task.
According to Lauriola and Levin (2001)’s results, people with low neuroticism and
people with high openness to experience took more risk in a decision making task,
while the correlation coefficients were not significant between risk-taking and other
personality traits (extraversion, conscientiousness, agreeableness). However, in our
experiment more agreeable people do more trials than others to be sure for the
result, they are risk-avoiders.
Hooper et al. (2008) and Byrne (2015) found that there was a negative relationship
between neuroticism and decision-making task performance. In our study,
neuroticism trait does not have any correlation with IST performance, but less
conscientious people show better performance.
100
Table 48: Test Summary of All Personality Traits for IST Performance
Measures
IST 1 -
Performance
IST 2 -
Performance IST1 Time IST2 Time
IST1
TrialNo
IST2
TrialNo
Pearson Correlation -.239 .103 .088 .013 -.082 -.202
Sig. (2-tailed) .154 .542 .606 .941 .631 .230
N 37 37 37 37 37 37
Pearson Correlation -,398* .123 .228 .155 .026 -.148
Sig. (2-tailed) .015 .469 .174 .360 .878 .381
N 37 37 37 37 37 37
Pearson Correlation -.084 .148 -.286 -.190 -,341* -.153
Sig. (2-tailed) .619 .383 .086 .259 .039 .364
N 37 37 37 37 37 37
Pearson Correlation .122 .022 .202 .084 .156 .074
Sig. (2-tailed) .471 .895 .231 .622 .356 .662
N 37 37 37 37 37 37
Pearson Correlation -.027 -.042 -.141 -.098 -.088 -.142
Sig. (2-tailed) .873 .807 .405 .566 .603 .400
N 37 37 37 37 37 37
Pearson Correlation -.081 -.063 .062 -.078 .041 -.017
Sig. (2-tailed) .633 .710 .717 .647 .810 .920
N 37 37 37 37 37 37
PF5 (Openness to
Experience)
PF6 (Negative Valence)
PF1 (Extraversion)
PF2 (Conscientiousness)
PF3 (Agreableness)
PF4 (Neuroticism)
101
Table 49: Group Statistics of All Personality Traits for IST Performance
Measures
Personality Trait Measurement N MeanStd.
Deviation
Std. Error
Mean
lower 16 1031.25 177.834 44.459
higher 21 952.38 227.198 49.579
lower 16 153872 66390.64 16597.7
higher 21 165839 58657.71 12800.2
lower 16 171.75 54.773 13.693
higher 21 166.76 54.176 11.822
lower 16 1081.88 328.638 82.159
higher 21 936.19 340.786 74.365
lower 16 162058 69141.4 17285.4
higher 21 159882 71733.1 15653.4
lower 16 137.75 40.575 10.144
higher 21 131.14 44.296 9.666
lower 24 1016.67 194.862 39.776
higher 13 930.77 228.709 63.432
lower 24 149994 57797.32 11797.8
higher 13 180363 65588.17 18190.9
lower 24 170.83 55.645 11.359
higher 13 165.38 52.01 14.425
lower 24 986.67 326.132 66.571
higher 13 1022.31 373.991 103.726
lower 24 159979 69212.15 14127.9
higher 13 162381 73278.99 20323.9
lower 24 139.83 41.281 8.427
higher 13 123.23 43.603 12.093
lower 20 965 203.328 45.465
higher 17 1011.76 217.607 52.778
lower 20 172745 69119.88 15455.7
higher 17 146452 49437.06 11990.2
lower 20 180.6 50.983 11.4
higher 17 155.18 55.115 13.367
lower 20 966 355.252 79.437
higher 17 1038.24 324.966 78.816
lower 20 177164 77678.02 17369.3
higher 17 141598 54997.91 13339
lower 20 136.15 45.383 10.148
higher 17 131.47 39.529 9.587
IST2 - Performance
IST2 - Time
IST2 - TrialNo
PF3 (Agreeableness)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
IST2 - Performance
IST2 - Time
IST2 - TrialNo
PF2 (Conscientiousness)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
IST2 - Performance
IST2 - Time
IST2 - TrialNo
PF1 (Extraversion)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
102
Table 49 (continued): Group Statistics of All Personality Traits for IST
Performance Measures
Personality Trait Measurement N MeanStd.
Deviation
Std. Error
Mean
lower 19 963.16 208.728 47.885
higher 18 1011.11 211.128 49.763
lower 19 150905 65525.56 15032.6
higher 18 170965 56996.58 13434.2
lower 19 168.42 57.909 13.285
higher 18 169.44 50.618 11.931
lower 19 968.42 366.14 83.998
higher 18 1031.67 314.872 74.216
lower 19 165155 79736.62 18292.8
higher 18 156251 59154.95 13943
lower 19 133.21 46.413 10.648
higher 18 134.83 38.756 9.135
lower 18 983.33 194.785 45.911
higher 19 989.47 225.819 51.806
lower 18 161640 67067.69 15808
higher 19 159740 57605.05 13215.5
lower 18 170.94 52.275 12.321
higher 19 167 56.436 12.947
lower 18 1034.44 347.431 81.89
higher 19 965.79 336.639 77.23
lower 18 160047 67128.84 15822.4
higher 19 161558 73799.97 16930.9
lower 18 131.61 35.119 8.278
higher 19 136.26 48.971 11.235
lower 27 1018.52 184.051 35.421
higher 10 900 253.859 80.277
lower 27 158194 62253.86 11980.8
higher 10 167334 62233.41 19679.9
lower 27 171.04 54.852 10.556
higher 10 163.2 52.971 16.751
lower 27 989.26 348.137 66.999
higher 10 1026 329.046 104.053
lower 27 163940 69501.1 13375.5
higher 10 152407 73107.72 23118.7
lower 27 136.63 39.783 7.656
higher 10 126.9 49.992 15.809
PF6 (Negative Valence)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
IST2 - Performance
IST2 - Time
IST2 - TrialNo
PF5 (Openness to Experience)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
IST2 - Performance
IST2 - Time
IST2 - TrialNo
PF4 (Neuroticism)
IST1 - Performance
IST1 - Time
IST1 - TrialNo
IST2 - Performance
IST2 - Time
IST2 - TrialNo
103
Lastly, Pearson correlation coefficient was calculated to find out whether there is a
correlation between n-back performances and personality traits. Results show that it
cannot be said that there is a correlation between any personality trait and any n-
back performance measures since all p values are higher than .05. Test summary and
group statistics of each personality trait for each n-back performance measure can be
found in Table 50 and Table 51, respectively.
Table 50: Test Summary of All Personality Traits for N-back Performance
Measures
back1 back2 back3 back4
Pearson Correlation -.027 .094 -.082 -.119
Sig. (2-tailed) .875 .580 .630 .481
N 37 37 37 37
Pearson Correlation .048 .094 .133 -.293
Sig. (2-tailed) .778 .582 .431 .079
N 37 37 37 37
Pearson Correlation .294 .126 -.037 .103
Sig. (2-tailed) .078 .456 .829 .543
N 37 37 37 37
Pearson Correlation -.063 .128 -.041 -.137
Sig. (2-tailed) .713 .450 .808 .418
N 37 37 37 37
Pearson Correlation .122 .122 .020 .136
Sig. (2-tailed) .473 .472 .904 .421
N 37 37 37 37
Pearson Correlation .323 .190 -.044 .248
Sig. (2-tailed) .051 .261 .795 .139
N 37 37 37 37
PF4 (Neuroticism)
PF5 (Openness to
Experience)
PF6 (Negative Valence)
PF1 (Extraversion)
PF2 (Conscientiousness)
PF3 (Agreableness)
104
Table 51: Group Statistics of All Personality Traits for N-back Performance
Measures
Personality Trait N MeanStd.
Deviation
Std.
Error
Mean
lower 16 93.19 12.608 3.152
higher 21 91.05 14.03 3.062
lower 16 76.56 22.085 5.521
higher 21 78.19 20.665 4.51
lower 16 44.69 21.039 5.26
higher 21 40.95 18.906 4.126
lower 16 34.13 18.04 4.51
higher 21 30.1 16.891 3.686
lower 24 91.13 13.911 2.839
higher 13 93.54 12.461 3.456
lower 24 74.63 22.604 4.614
higher 13 82.77 17.249 4.784
lower 24 41.38 16.5 3.368
higher 13 44.77 25.094 6.96
lower 24 31.83 17.643 3.601
higher 13 31.85 17.262 4.788
lower 20 90.65 14.091 3.151
higher 17 93.53 12.536 3.04
lower 20 76.15 21.495 4.806
higher 17 79.06 20.954 5.082
lower 20 43.7 19.618 4.387
higher 17 41.24 20.228 4.906
lower 20 30.55 17.913 4.006
higher 17 33.35 16.893 4.097
4-back
1-back
2-back
3-back
4-back
1-back
2-back
3-back
4-back
1-back
2-back
3-back
Measurement
PF1 (Extraversion)
PF2 (Conscientiousness)
PF3 (Agreeableness)
105
Table 51 (continued): Group Statistics of All Personality Traits for N-back
Performance Measures
Personality Trait N MeanStd.
Deviation
Std.
Error
Mean
lower 19 92 12.953 2.972
higher 18 91.94 14.023 3.305
lower 19 74.84 23.573 5.408
higher 18 80.28 18.159 4.28
lower 19 44.53 20.048 4.599
higher 18 40.5 19.6 4.62
lower 19 35.11 16.656 3.821
higher 18 28.39 17.697 4.171
lower 18 89.61 14.508 3.42
higher 19 94.21 11.993 2.751
lower 18 74.94 21.74 5.124
higher 19 79.89 20.575 4.72
lower 18 42 20.059 4.728
higher 19 43.11 19.81 4.545
lower 18 29.67 18.124 4.272
higher 19 33.89 16.643 3.818
lower 27 90.07 14.204 2.734
higher 10 97.1 9.171 2.9
lower 27 76.67 23.339 4.492
higher 10 79.7 13.549 4.284
lower 27 44.81 18.614 3.582
higher 10 36.5 22.117 6.994
lower 27 30.74 14.203 2.733
higher 10 34.8 24.426 7.724
Measurement
2-back
3-back
4-back
1-back
2-back
3-back
4-back
1-back
2-back
PF4 (Neuroticism)
PF5 (Openness to
Experience)
PF6 (Negative Valence)
3-back
4-back
1-back
106
Gray (2001) investigated extraversion and neuroticism to see whether they had any
relationship with verbal and spatial 2-back tasks and found out that these traits did
not have any correlations. Even though we cannot compare our study for further n-
backs like 3 and 4 and with other traits (openness, conscientiousness and
agreeableness), it can be said that our finding is similar with that of Gray (2001).
4.10. Personality Differences in fNIR Measures
As we mention in earlier part, there is fNIR data of 19 participants out of 37 due to
physical problems. Brain activity information about left DLPFC by voxel-1 and
right DLPFC by voxel-4 were used.
Pearson correlation coefficient was conducted to see whether there is a correlation
between personality traits (extraversion, conscientiousness, agreeableness,
neuroticism, openness to experience and negative valence) and fNIR measures
(oxygenation level in voxel-1 (left DLPFC), in voxel-4 (right DLPFC)) across all 19
participants.
First, the correlation was conducted for measures in voxel-1. Results show that
conscientiousness is negatively correlated with oxygenation level in voxel-1 during
maze task, r(19)= -.548, p≤.05. Extraversion (r(19)= .459, p≤.05), agreeableness
(r(19)= .483, p≤.05), and openness to experience (r(19)= .518, p≤.05) are all
positively correlated with oxygenation level in voxel-1 during ist-2. Openness to
experience also correlates positively with oxygenation level in voxel-1 during 2-
back (r(19)= .556, p≤.05) and 3-back (r(19)= .553, p≤.05).
Correlations can be seen in Table 52. Group statistics of each personality trait
(extraversion, conscientiousness, agreeableness, neuroticism, openness to experience
and negative valence) for fNIR measures are given in Table 54, 55, 56, 57, 58 and
59, respectively.
107
Table 52: Test Summary of All Personality Traits for fNIR Measures of Left
DLPFC
For the relationship between oxygenation level in voxel-4 and personality traits,
Pearson correlation coefficient was conducted and results show that agreeableness is
positively correlated with oxygenation level in voxel-4 during ist-1, r(19)= .560,
p≤.05. Openness to experience show positive correlations with oxygenation level in
voxel-4 during 2-back, r(19)= .598, p≤.05 and 3-back, r(19)= .609, p≤.05 and 4-
back, r(19)= .534, p≤.05.
Correlations can be seen in Table 53. Group statistics of each personality trait
(extraversion, conscientiousness, agreeableness, neuroticism, openness to experience
and negative valence) for fNIR measures are given in Table 54, 55, 56, 57, 58 and
59, respectively.
V1_maze V1_ist1 V1_ist2 V1_1back V1_2back V1_3back V1_4back
Pearson Correlation -.123 .359 ,459* .035 .206 .111 .154
Sig. (2-tailed) .617 .131 .048 .891 .411 .662 .541
N 19 19 19 19 19 19 19
Pearson Correlation -,548* -.404 -.268 .034 -.117 -.188 .075
Sig. (2-tailed) .015 .087 .268 .895 .645 .455 .768
N 19 19 19 19 19 19 19
Pearson Correlation .400 .370 ,483* .203 .407 .393 .325
Sig. (2-tailed) .090 .119 .036 .420 .093 .106 .188
N 19 19 19 19 19 19 19
Pearson Correlation -.214 -.090 -.383 .116 -.184 -.348 -.147
Sig. (2-tailed) .380 .715 .105 .646 .464 .157 .560
N 19 19 19 19 19 19 19
Pearson Correlation .099 .109 ,518* .103 ,556
*,553
* .399
Sig. (2-tailed) .688 .657 .023 .685 .017 .017 .101
N 19 19 19 19 19 19 19
Pearson Correlation .310 -.078 -.024 .235 .071 -.052 -.026
Sig. (2-tailed) .197 .751 .923 .347 .779 .838 .917
N 19 19 19 19 19 19 19
PF5 (Openness to
Experience)
PF6 (Negative Valence)
PF1 (Extraversion)
PF2
(Conscientiousness)
PF3 (Agreableness)
PF4 (Neuroticism)
108
Table 53: Test Summary of All Personality Traits for fNIR Measures of Right
DLPFC
V4_maze V4_ist1 V4_ist2 V4_1back V4_2back V4_3back V4_4back
Pearson Correlation .076 .406 .454 .086 .189 .101 .050
Sig. (2-tailed) .758 .085 .051 .735 .453 .690 .844
N 19 19 19 18 18 18 18
Pearson Correlation .002 .000 -.051 .197 .153 -.005 .125
Sig. (2-tailed) .994 .999 .834 .432 .543 .983 .620
N 19 19 19 18 18 18 18
Pearson Correlation .335 ,560* .411 .317 .265 .390 .402
Sig. (2-tailed) .161 .013 .081 .200 .289 .110 .098
N 19 19 19 18 18 18 18
Pearson Correlation -.319 -.341 -.122 -.113 -.331 -.410 -.427
Sig. (2-tailed) .183 .153 .620 .656 .180 .091 .077
N 19 19 19 18 18 18 18
Pearson Correlation .160 .375 .358 .325 ,598**
,609**
,534*
Sig. (2-tailed) .513 .113 .133 .189 .009 .007 .022
N 19 19 19 18 18 18 18
Pearson Correlation -.285 -.249 -.191 .035 .081 .061 -.019
Sig. (2-tailed) .237 .305 .432 .890 .749 .811 .939
N 19 19 19 18 18 18 18
PF4 (Neuroticism)
PF5 (Openness to
Experience)
PF6 (Negative Valence)
PF1 (Extraversion)
PF2
(Conscientiousness)
PF3 (Agreableness)
109
Table 54: Group Statistics of Extraversion Personality Trait for fNIR Measures
N MeanStd.
Deviation
Std.
Error
Mean
lower 6 1.004 0.995 0.406
higher 13 0.409 1.035 0.287
lower 6 0.343 1.029 0.42
higher 13 0.507 0.843 0.234
lower 6 -0.122 0.768 0.313
higher 13 0.741 0.827 0.229
lower 6 0.095 0.79 0.323
higher 13 0.754 1.052 0.292
lower 6 0.587 0.751 0.306
higher 13 1.45 0.898 0.249
lower 6 0.689 0.922 0.376
higher 13 1.582 0.991 0.275
lower 6 0.109 1.265 0.516
higher 13 0.05 1.064 0.295
lower 6 0.216 1.074 0.439
higher 13 0.162 0.85 0.236
lower 6 0.802 0.789 0.322
higher 13 1.061 1.392 0.386
lower 6 0.456 1.596 0.651
higher 13 0.982 1.242 0.344
lower 6 1.529 1.025 0.418
higher 13 1.563 1.515 0.42
lower 6 1.203 1.699 0.693
higher 13 1.217 1.381 0.383
lower 6 1.701 2.059 0.841
higher 13 1.771 1.781 0.494
lower 6 1.452 1.795 0.733
higher 13 1.554 1.692 0.469
V1_4back
V4_4back
V1_1back
V4_1back
V1_2back
V4_2back
V1_3back
V4_3back
V1_maze
V4_maze
V1_ist1
V4_ist1
V1_ist2
V4_ist2
Extraversion
110
Table 55: Group Statistics of Conscientiousness Personality Trait for fNIR
Measures
N MeanStd.
Deviation
Std.
Error
Mean
lower 13 0.74 1.074 0.298
higher 6 0.287 0.957 0.391
lower 13 0.504 0.926 0.257
higher 6 0.349 0.842 0.344
lower 13 0.466 0.541 0.15
higher 6 0.475 1.463 0.597
lower 13 0.4 0.923 0.256
higher 6 0.861 1.19 0.486
lower 13 1.261 0.723 0.201
higher 6 0.998 1.34 0.547
lower 13 1.262 0.755 0.21
higher 6 1.384 1.574 0.643
lower 13 -0.035 0.947 0.263
higher 6 0.294 1.441 0.588
lower 13 0.041 0.739 0.205
higher 6 0.479 1.193 0.487
lower 13 0.884 0.757 0.21
higher 6 1.186 1.97 0.804
lower 13 0.63 1.25 0.347
higher 6 1.22 1.561 0.637
lower 13 1.548 0.859 0.238
higher 6 1.562 2.188 0.893
lower 13 1.098 1.148 0.318
higher 6 1.461 2.047 0.836
lower 13 1.6 1.391 0.386
higher 6 2.072 2.652 1.083
lower 13 1.316 1.272 0.353
higher 6 1.968 2.423 0.989
Conscientiousness
V1_maze
V4_maze
V1_ist1
V4_ist1
V1_ist2
V4_ist2
V1_1back
V4_1back
V1_2back
V4_2back
V1_3back
V4_3back
V1_4back
V4_4back
111
Table 56: Group Statistics of Agreeableness Personality Trait for fNIR
Measures
N MeanStd.
Deviation
Std.
Error
Mean
lower 11 0.266 1.018 0.307
higher 8 1.053 0.927 0.328
lower 11 0.083 0.678 0.205
higher 8 0.967 0.902 0.319
lower 11 0.222 0.841 0.254
higher 8 0.809 0.889 0.314
lower 11 0.2 0.836 0.252
higher 8 1.021 1.073 0.379
lower 11 1.019 1.026 0.309
higher 8 1.396 0.789 0.279
lower 11 1.182 0.981 0.296
higher 8 1.463 1.154 0.408
lower 11 0.096 0.941 0.284
higher 8 0.031 1.348 0.477
lower 11 0.1 0.881 0.266
higher 8 0.288 0.966 0.342
lower 11 0.671 1.133 0.342
higher 8 1.404 1.272 0.45
lower 11 0.629 1.591 0.48
higher 8 1.074 0.936 0.331
lower 11 1.236 1.411 0.425
higher 8 1.988 1.211 0.428
lower 11 0.921 1.679 0.506
higher 8 1.613 0.993 0.351
lower 11 1.486 1.724 0.52
higher 8 2.111 1.992 0.704
lower 11 1.174 1.785 0.538
higher 8 2 1.488 0.526
Agreeableness
V1_maze
V4_maze
V1_ist1
V4_ist1
V1_ist2
V4_ist2
V1_1back
V4_1back
V1_2back
V4_2back
V1_3back
V4_3back
V1_4back
V4_4back
112
Table 57: Group Statistics of Neuroticism Personality Trait for fNIR Measures
N MeanStd.
Deviation
Std.
Error
Mean
lower 7 0.952 0.727 0.275
higher 12 0.39 1.155 0.334
lower 7 0.812 0.814 0.308
higher 12 0.247 0.881 0.254
lower 7 0.623 1.203 0.455
higher 12 0.379 0.692 0.2
lower 7 0.732 0.839 0.317
higher 12 0.437 1.111 0.321
lower 7 1.537 1.107 0.418
higher 12 0.968 0.784 0.226
lower 7 1.366 1.216 0.46
higher 12 1.262 0.971 0.28
lower 7 -0.15 1.221 0.461
higher 12 0.196 1.05 0.303
lower 7 0.215 1.216 0.459
higher 12 0.158 0.712 0.206
lower 7 1.108 0.926 0.35
higher 12 0.905 1.392 0.402
lower 7 1.185 1.176 0.445
higher 12 0.601 1.432 0.414
lower 7 1.899 0.999 0.378
higher 12 1.35 1.522 0.439
lower 7 1.653 1.004 0.38
higher 12 0.956 1.627 0.47
lower 7 1.823 1.892 0.715
higher 12 1.706 1.853 0.535
lower 7 1.853 1.188 0.449
higher 12 1.329 1.926 0.556
Neuroticism
V1_maze
V4_maze
V1_ist1
V4_ist1
V1_ist2
V4_ist2
V1_1back
V4_1back
V1_2back
V4_2back
V1_3back
V4_3back
V1_4back
V4_4back
113
Table 58: Group Statistics of Openness to Experience Personality Trait for
fNIR Measures
N Mean
Std.
Deviation
Std. Error
Mean
lower 8 .329 1.077 .381
higher 11 .792 1.007 .304
lower 8 .119 .919 .325
higher 11 .699 .803 .242
lower 8 .318 .581 .205
higher 11 .579 1.074 .324
lower 8 .111 .952 .337
higher 11 .862 .960 .289
lower 8 .726 .883 .312
higher 11 1.506 .852 .257
lower 8 1.011 .907 .321
higher 11 1.511 1.113 .336
lower 8 .104 .929 .328
higher 11 .043 1.247 .376
lower 8 -.021 .570 .202
higher 11 .325 1.079 .325
lower 8 .504 .932 .329
higher 11 1.325 1.321 .398
lower 8 .081 1.145 .405
higher 11 1.351 1.254 .378
lower 8 1.021 1.399 .495
higher 11 1.938 1.233 .372
lower 8 .451 1.414 .500
higher 11 1.767 1.237 .373
lower 8 1.228 1.688 .597
higher 11 2.128 1.887 .569
lower 8 .784 1.497 .529
higher 11 2.059 1.652 .498
V4_4back
V4_1back
V1_2back
V4_2back
V1_3back
V4_3back
V1_4back
V4_maze
V1_ist1
V4_ist1
V1_ist2
V4_ist2
V1_1back
Openness to Experience
V1_maze
114
Table 59: Group Statistics of Negative Valence Personality Trait for fNIR
Measures
N Mean
Std.
Deviation
Std. Error
Mean
lower 13 .510 1.099 .305
higher 6 .787 .942 .385
lower 13 .694 .886 .246
higher 6 -.063 .656 .268
lower 13 .569 .860 .239
higher 6 .253 .991 .405
lower 13 .746 1.051 .291
higher 6 .113 .811 .331
lower 13 1.239 .895 .248
higher 6 1.045 1.075 .439
lower 13 1.478 1.082 .300
higher 6 .915 .889 .363
lower 13 -.045 1.231 .341
higher 6 .316 .761 .311
lower 13 .042 .902 .250
higher 6 .478 .885 .361
lower 13 .974 1.370 .380
higher 6 .992 .909 .371
lower 13 .611 1.033 .286
higher 6 1.262 1.888 .771
lower 13 1.599 1.522 .422
higher 6 1.450 .992 .405
lower 13 1.147 1.398 .388
higher 6 1.356 1.653 .675
lower 13 1.758 2.009 .557
higher 6 1.730 1.473 .601
lower 13 1.517 1.708 .474
higher 6 1.533 1.761 .719
V1_2back
V4_2back
V1_3back
V4_3back
V1_4back
V4_4back
V1_ist1
V4_ist1
V1_ist2
V4_ist2
V1_1back
V4_1back
Negative Valence
V1_maze
V4_maze
115
In the literature, there are very few studies on personality differences by
neuroimaging. Canli et al. (2001) investigated the Big Five personality traits which
had five personality dimensions by FMRI but using only women participants. They
found out that brain activation was positively correlated with extraversion and
negatively correlated with neuroticism.
Our findings show extraversion has a relationship with brain activations during IST-
2; conscientiousness has a relationship with brain activations during maze,
agreeableness has a relationship with brain activations during IST-1 and IST-2;
openness to experience has a relationship with brain activations during IST-2, 2-
back, 3-back and 4-back. However neuroticism and negative valence has no
relationship with blood oxygenation level during tasks. Lack of studies on that topic
makes it hard to interpret and compare our findings with literature.
116
117
CHAPTER 5
CONCLUSION
5.1. Outcomes of the Study
In this study, gender and personality differences on cognitive tasks are studied. Two
types of measures are used in study. One of them is called performance measure and
it depends on performance of participants, and fNIR device is used as a second
measure to assess human mental workload.
There are three different cognitive tasks used in the study: N-back as a working-
memory task, IST as a decision making task, maze as a spatial task. N-back has four,
IST has two difficulty levels and maze has one difficulty level.
No gender difference in maze performance measures exists even though literature
indicates males are significantly better than females. Lack of variety in participants
and the simplicity of maze may cause this disparity. We can conclude that gender is
not a criterion for spatial based tasks such as a maze task, if people have similar
background according to our results. Different results can be found with participants
with different specialty and/or a more complex maze task.
Gender difference on trial numbers during IST-2 is expected according to literature.
It is observed that females make their choices according to their instincts, whereas
males analyze data and make their choice according to facts. No gender differences
do not come up for other IST performance measures. Gender do not have any effects
on performance or completion time of a decision making task. We see clearly that
118
gender is not important, if decision making performance is cared. However, if the
number of trials is an essential consideration, gender should be looked into.
Males perform better than females during 4-back task, while there is no gender
difference during 1, 2, and 3-back tasks. Since males are better at spatial-based
working memory tasks according to literature (Coluccia & Louse, 2004; Galeo &
Kimura, 1993; Persson, 2013), it is an expected result. Gender difference was
observed at 2 or 3-back level on previous studies, but it runs across at 4th
level in our
study. Similar educational level of participants may prevent gender differences at 2
and 3-backs. We can use gender as a personnel selection criterion for jobs require
working memory abilities but educational and analytical background should also be
taken into consideration.
fNIR results show that there is no difference in oxygenation level at both left and
right DLPFC regions between men and women during all cognitive tasks. In other
words, gender does not have any effects on mental workload during a cognitive task.
This finding can be used in employment selections. If a job involves cognitive
complexity which can create mental workload, gender is not a criterion during a
personnel selection to this job for people with similar educational background.
Performance at different difficulty levels are investigated and people show less
accuracy at more difficult levels which is expected according to literature. Results of
fNIR tests show that rise in difficulty level of the task increases changes in
oxygenation level in voxels. This finding is expected and consistent with literature.
When participants work harder on the task at more difficult level, increase in
oxygenation level is observed as a symptom for human mental workload.
Results of our study show that for maze task, there is not a significant personality
difference in performance measures for each personality trait which is a similar
finding with literature. We conclude people with different personalities do not show
different performance during a spatial task like a maze task. Tasks which require
119
spatial abilities can be performed by any person with different personalities.
Personality does not affect the performance of the person. Therefore, spatial ability
based jobs should find another criteria than personality test for its personnel
selection.
Decision making task shows that less conscientious people show better performance
during IST-1 task. People with more conscientiousness have more performance
concerns and their competitive structure may cause poor performance at the start of
the task. Therefore, they show worse performance at the first part of the decision
making task, but they put themselves together at the second part of the task. Results
also show that more agreeable people do fewer trials than others. People with more
agreeableness care how they seem than how they actually are. They pay relatively
more importance to others’ thoughts. Since experiments were done under
observations, people with more agreeableness did fewer trials during IST-1 task to
show the observer that they completed their tasks fast to make observers to believe
that they were good at this. Only two personality traits- agreeableness and
conscientiousness- have a relationship with the decision making task during our
experiments. These results show us personality can affect the decision making task
performance and behavior of a person. During a personnel selection for a job based
on decision making tasks, personality traits can be used as a part of the job interview
process.
Results for personality differences of n-back tasks are similar with maze task. There
is no significant effect of personality differences in n-back performance measures.
We can conclude that different personalities do not affect short-memory task
performance. If a job requires a short-memory cognitive ability, the selection of a
person should be based on his/ her working side rather than his/her personality
factors.
Oxygenation level differences for each personality trait are searched and it is found
120
that oxygenation levels do not differ during maze and 1-back tasks for all
personality traits. Since task difficulties of 1-back and maze tasks are lower than
other cognitive tasks, no oxygenation level difference is expected. Our findings
show that extraversion has a positive correlation with brain activations during IST-2;
agreeableness has a positive correlation with brain activations during IST-1 and IST-
2. More extraverted and/or more agreeable people mind other people’s thoughts
about them. Thinking about thoughts of the observer during experiments may cause
the increase in their mental workload therefore they may result incremental changes
in the level of blood oxygenation.
Results show that conscientiousness has a negative correlation with brain activations
during the maze task. In other words, people with more conscientiousness show less
mental workload during the maze task. More conscientious people are more focused
on their success, so they are more oriented to tasks, ignore distractions. Handling a
task more consciously may provide less mental workload to people.
According to our findings, openness to experience has a positive correlation with
brain activations during IST-2, 2-back, 3-back and 4-back. More open people like to
try new things and they can be enthusiastic during doing those tasks. Our
experiments with all different tasks and tools may seem new and different to other
people. Therefore, more open people may be excited and alerted during the
experiment and that can cause the increase in blood oxygenation level. As seen
from results, the oxygenation level in a brain has a relationship with personality.
That shows how a personality is important and distinguishing on mental workload
during doing a job. Therefore, personality traits should be involved in the personnel
selection to the jobs especially to mentally challenging ones.
5.2. Limitations of the Study
In this study, even though 37 participants are recruited, data of 19 participants were
available to be used after filtering functions of fNIR. That creates loss of almost
121
%50 of the data. Therefore, experiments part of our study lasted more than expected.
Increased number of participants may give chance to reach more accurate findings.
Since three types of tasks are decided to be used, task times are short. Gender
differences might have been observed, if a more difficult and time-consuming maze
had been used.
All participants are undergraduate or graduate engineering students from the Middle
East Technical University or Hacettepe University. Such a narrow area of participant
structure can affect the results. In the future, participants from different fields and
different schools should be chosen in order to obtain more reliable results.
122
REFERENCES
Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and interests:
evidence for overlapping traits. Psychological Bulletin, 121(2), 219.
Agor, W. H. (1986). The logic of intuitive decision making: A research-based
approach for top management. Quorum Books.
Ayaz, H., Allen, S.L., Platek, S.M., & Onaral, B. (2008). Maze Suite 1.0: a complete
set of tools to prepare, present, and analyze navigational and spatial cognitive
neuroscience experiments. Behavior Research Methods, 40 (1), 353-359.
Ayaz, H., Izzetoglu, M., Shewokis, P., & Onaral, B. (2010, August). Sliding-
window motion artifact rejection for functional near-infrared spectroscopy. In
Engineering in Medicine and Biology Society (EMBC), 2010 Annual International
Conference of the IEEE (pp. 6567-6570). IEEE.
Ayaz, H., Shewokis, P. A., Bunce, S., Izzetoglu, K., Willems, B., & Onaral, B.
(2012). Optical brain monitoring for operator training and mental workload
assessment. Neuroimage, 59(1), 36-47.
Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity
to future consequences following damage to human prefrontal cortex. Cognition,
50(1), 7-15.
Bell, E. C., Willson, M. C., Wilman, A. H., Dave, S., & Silverstone, P. H. (2006).
Males and females differ in brain activation during cognitive tasks. Neuroimage,
30(2), 529-538.
123
Blokland, G. A., McMahon, K. L., Hoffman, J., Zhu, G., Meredith, M., Martin, N.
G., Thompson, P.M., Zubicaray, G.I., & Wright, M. J. (2008). Quantifying the
heritability of task-related brain activation and performance during the N-back
working memory task: a twin fMRI study. Biological Psychology, 79(1), 70-79.
Boles, D.B. & Adair, L.P. (2001). The multiple resources questionnaire (MRQ).
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 45
(25), 1790-1794.
Bolla, K. I., Eldreth, D. A., Matochik, J. A., & Cadet, J. L. (2004). Sex-related
differences in a gambling task and its neurological correlates. Cerebral Cortex,
14(11), 1226-1232.
Borghini, G., Astolfi, L., Vecchiato, G., Mattia, D., & Babiloni, F. (2014).
Measuring neurophysiological signals in aircraft pilots and car drivers for the
assessment of mental workload, fatigue and drowsiness. Neuroscience &
Biobehavioral Reviews, 44, 58-75.
Borghini, G., Isabella, R., Vecchiato, G., Toppi, J., Astolfi, L., Caltagirone, C., &
Babiloni, F. (2011). Brainshield: HREEG study of perceived pilot mental workload.
Italian Journal of Aerospace Medicine, 5, 34-47
Braver, T. S., Barch, D. M., Gray, J. R., Molfese, D. L., & Snyder, A. (2001).
Anterior cingulate cortex and response conflict: effects of frequency, inhibition and
errors. Cerebral Cortex, 11(9), 825-836.
Brookings, J. B., Wilson, G. F., & Swain, C. R. (1996). Psychophysiological
responses to changes in workload during simulated air traffic control. Biological
Psychology, 42(3), 361-377.
Bunce, S. C., Izzetoglu, M., Izzetoglu, K., Onaral, B., & Pourrezaei, K. (2006).
Functional near-infrared spectroscopy. Engineering in Medicine and Biology
124
Magazine, IEEE, 25(4), 54-62.
Byrne, K. A., Silasi-Mansat, C. D., & Worthy, D. A. (2015). Who chokes under
pressure? The Big Five personality traits and decision-making under pressure.
Personality and Individual Differences, 74, 22-28.
Canli, T., Zhao, Z., Desmond, J. E., Kang, E., Gross, J., & Gabrieli, J. D. (2001). An
fMRI study of personality influences on brain reactivity to emotional stimuli.
Behavioral Neuroscience, 115(1), 33.
Cansiz, Y., 2012. Effects of way finding affordances on usability of virtual world
environments in terms of users’ satisfaction, performance, and mental workload:
examination by eye-tracking and fNIR device. MS. Thesis, Middle East Technical
University, Ankara, Turkey.
Carlson, S., Martinkauppi, S., Rämä, P., Salli, E., Korvenoja, A., & Aronen, H. J.
(1998). Distribution of cortical activation during visuospatial n-back tasks as
revealed by functional magnetic resonance imaging. Cerebral Cortex, 8(8), 743-752.
Case, R., Kurland, D. M., & Goldberg, J. (1982). Operational efficiency and the
growth of short-term memory span. Journal of Experimental Child Psychology,
33(3), 386-404.
Chamorro-Premuzic, T., Furnham, A., & Petrides, K. (2006). Personality and
intelligence. Journal of Individual Differences, 27(3), 147-150.
Chance, B., Anday, E., Nioka, S., Zhou, S., Hong, L., Worden, K., Li, C., Murray,
T., Ovetsky, Y., Pidikiti, D. & Thomas, R. (1998). A novel method for fast imaging
of brain function non-invasively with light. Optics Express, 2(10), 411-423.
Clark, L., Robbins, T. W., Ersche, K. D., & Sahakian, B. J. (2006). Reflection
impulsivity in current and former substance users. Biological Psychiatry, 60(5), 515-
522.
125
Colle, H. A., & Reid, G. B. (1999). Double trade-off curves with different cognitive
processing combinations: Testing the cancellation axiom of mental workload
measure theory. Human Factors: The Journal of the Human Factors and Ergonomics
Society, 41(1), 35-50.
Coluccia, E., & Louse, G. (2004). Gender differences in spatial orientation: A
review. Journal of Environmental Psychology, 24(3), 329-340.
Costa, P. T., Fozard, J. L., McCrae, R. R., & Bosse, R. (1976). Relations of age and
personality dimensions to cognitive ability factors. Journal of Gerontology, 31(6),
663-669.
Coyle, S. M., Ward, T. E., & Markham, C. M. (2007). Brain–computer interface
using a simplified functional near-infrared spectroscopy system. Journal of Neural
Engineering, 4(3), 219-226.
Crone, E. A., Vendel, I., & van der Molen, M. W. (2003). Decision-making in
disinhibited adolescents and adults: insensitivity to future consequences or driven by
immediate reward?. Personality and Individual Differences, 35(7), 1625-1641.
Cui, X., Bray, S., Bryant, D. M., Glover, G. H., & Reiss, A. L. (2011). A
quantitative comparison of NIRS and fMRI across multiple cognitive tasks.
Neuroimage, 54(4), 2808-2821.
Damos, D. L., & Bloem, K. A. (1985). Type A behavior pattern, multiple-task
performance, and subjective estimation of mental workload. Bulletin of the
Psychonomic Society, 23(1), 53-56.
Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory
and reading. Journal of Verbal Learning and Verbal Behavior, 19(4), 450-466.
Davies, A. D. (1965). The perceptual maze test in a normal population. Perceptual
and Motor Skills, 20(1), 287-293.
126
Deco, G., & Rolls, E. T. (2005). Attention, short-term memory, and action selection:
a unifying theory. Progress in Neurobiology, 76(4), 236-256.
DeYoung, C. G. (2010). Personality neuroscience and the biology of traits. Social
and Personality Psychology Compass, 4(12), 1165-1180.
De Rivecourt, M., Kuperus, M. N., Post, W. J., & Mulder, L. J. M. (2008).
Cardiovascular and eye activity measures as indices for momentary changes in
mental effort during simulated flight. Ergonomics, 51(9), 1295-1319.
De Waard, D. (1996). The measure of drivers’ mental workload. PhD thesis,
University of Groningen, The Netherlands.
Dussault, C., Jouanin, J. C., Philippe, M., & Guezennec, C. Y. (2005). EEG and
ECG changes during simulator operation reflect mental workload and vigilance.
Aviation, Space, and Environmental Medicine, 76(4), 344-351.
Eggemeier, F.T. (1988). Properties of workload assessment techniques. In: Hancock,
P.A., Meshkati, N. (Eds.), Human Mental Workload. North-Holland, Amsterdam,
41-62.
Field, A. (2009). Discovering statistics using SPSS. Sage publications.
Friedman, M., & Rosenman, R. H. (1974). Type A behavior and your heart. New
York: Knopf.
Fujimoto, T., Matsumoto, T., Fujita, S., Takeuchi, K., Nakamura, K., Mitsuyama,
Y., & Kato, N. (2008). Changes in glucose metabolism due to aging and gender-
related differences in the healthy human brain. Psychiatry Research: Neuroimaging,
164(1), 58-72.
Furnham, A., & Chamorro-Premuzic, T. (2006). Personality, intelligence and
general knowledge. Learning and Individual Differences, 16(1), 79-90.
127
Fuster, J. M. (2008). The prefrontal cortex. 4. China: Elsevier.
Galea, L. A., & Kimura, D. (1993). Sex differences in route-learning. Personality
and Individual Differences, 14(1), 53-65.
Galley, N. (1993). The evaluation of the electrooculogram as a psychophysiological
measuring instrument in the driver study of driver behaviour. Ergonomics, 36(9),
1063-1070.
Gencoz, T., & Oncul, O. (2012). Examination of Personality Characteristics in a
Turkish Sample: Development of Basic Personality Traits Inventory. The Journal of
General Psychology, 139(3), 194-216.
Gevins, A., Smith, M. E., Le, J., Leong, H., Bennett, J., Martin, N., McEvoy, L., Du,
R. & Whitfield, S. (1996). High resolution evoked potential imaging of the cortical
dynamics of human working memory. Electroencephalography and Clinical
Neurophysiology, 98(4), 327-348.
Gopher, D. & Braune, R. (1984). On the psychophysics of workload-why bother
with subjective measures. Human Factors, 26(5), 519–532.
Graham, E. K., & Lachman, M. E. (2012). Personality stability is associated with
better cognitive performance in adulthood: Are the stable more able?. The Journals
of Gerontology Series B: Psychological Sciences and Social Sciences, 67(5), 545-
554.
Graham, E. K., & Lachman, M. E. (2014). Personality traits, facets and cognitive
performance: Age differences in their relations. Personality and Individual
Differences, 59, 89-95.
Gray, J. R. (2001). Emotional modulation of cognitive control: Approach–
withdrawal states double-dissociate spatial from verbal two-back task performance.
Journal of Experimental Psychology: General, 130(3), 436.
128
Greely, H. T., & Wagner, A. D. (2011). Reference guide on neuroscience: Third
Edition. Federal judicial center reference manual on scientific evidence, Federal
Judicial Center, Washington, DC, 747-812.
Gron, G., Wunderlich, A. P., Spitzer, M., Tomczak, R., & Riepe, M. W. (2000).
Brain activation during human navigation: gender-different neural networks as
substrate of performance. Nature Neuroscience, 3(4), 404-408.
Gronwall, D. M. A. (1977). Paced auditory serial-addition task: a measure of
recovery from concussion. Perceptual and Motor Skills, 44(2), 367-373.
Hankins, T. C., & Wilson, G. F. (1998). A comparison of heart rate, eye activity,
EEG and subjective measures of pilot mental workload during flight. Aviation Space
and Environmental Medicine, 69(4), 360-367.
Hart, S.G. & Staveland, L.E. (1988). Development of NASA-TLX (Task Load
Index): Results of empirical and theoretical research. Advances in Psychology, 52,
139-183.
Harvey, P. O., Fossati, P., Pochon, J. B., Levy, R., LeBastard, G., Lehericy, S.,
Allilaire, J.F., & Dubois, B. (2005). Cognitive control and brain resources in major
depression: an fMRI study using the n-back task. Neuroimage, 26(3), 860-869.
Haut, K. M., & Barch, D. M. (2006). Sex influences on material-sensitive functional
lateralization in working and episodic memory: men and women are not all that
different. Neuroimage, 32(1), 411-422.
Hooper, C. J., Luciana, M., Wahlstrom, D., Conklin, H. M., & Yarger, R. S. (2008).
Personality correlates of Iowa Gambling Task performance in healthy adolescents.
Personality and Individual Differences, 44(3), 598-609.
Izzetoglu, K., Ayaz, H., Merzagora, A., Izzetoglu, M., Shewokis, P. A., Bunce, S.
C., Pourrezaei, K., Rosen, A. & Onaral, B. (2011). The evolution of field deployable
129
fNIR spectroscopy from bench to clinical settings. Journal of Innovative Optical
Health Sciences, 4(03), 239-250.
Izzetoglu, K., Bunce, S., Onaral, B., Pourrezaei, K., & Chance, B. (2004).
Functional optical brain imaging using near-infrared during cognitive tasks.
International Journal of Human-Computer Interaction, 17(2), 211-227.
Jex, H.R. (1988). Measuring mental workload: Problems, progress, and promises In:
Hancock, P.A., Meshkati, N. (Eds.), Human Mental Workload. North-Holland,
Amsterdam, 5-39.
Johnson, P. (1992). Human computer interaction: psychology, task analysis, and
software engineering. McGraw-Hill.
Jorna, P. G. A. M. (1993). Heart-rate and workload variations in actual and
simulated flight. Ergonomics, 36(9), 1043-1054.
Judge, T. A., Higgins, C. A., Thoresen, C. J., & Barrick, M. R. (1999). The big five
personality traits, general mental ability, and career success across the life span.
Personnel Psychology, 52(3), 621-652.
Koch, K., Pauly, K., Kellermann, T., Seiferth, N. Y., Reske, M., Backes, V.,
Stocker, T., Shah, J.N., Amunts, K., Kircher, T., Schneider, F., & Habel, U. (2007).
Gender differences in the cognitive control of emotion: An fMRI study.
Neuropsychologia, 45(12), 2744-2754
Lee, D. H., Parks, K.S. (1990). Multivariate analysis of mental and physical load
components in sinus arrhythmia scores. Ergonomics, 33(1), 35-47.
León-Carrión, J., Izzetoglu, M., Izzetoglu, K., Martín-Rodríguez, J. F., Damas-
López, J., Martin, J. M. B., & Domínguez-Morales, M. R. (2010). Efficient learning
produces spontaneous neural repetition suppression in prefrontal cortex. Behavioural
Brain Research, 208(2), 502-508.
130
Lauriola, M., & Levin, I. P. (2001). Personality traits and risky decision-making in a
controlled experimental task: An exploratory study. Personality and Individual
Differences, 31(2), 215-226.
Li, T., Luo, Q., & Gong, H. (2010). Gender-specific hemodynamics in prefrontal
cortex during a verbal working memory task by near-infrared spectroscopy.
Behavioural Brain Research, 209(1), 148-153.
Lighthall, N. R., Sakaki, M., Vasunilashorn, S., Nga, L., Somayajula, S., Chen, E.
Y., Samii, N. & Mather, M. (2012). Gender differences in reward-related decision
processing under stress. Social Cognitive and Affective Neuroscience, 7(4), 476-
484.
Mandrick, K., Derosiere, G., Dray, G., Coulon, D., Micallef, J. P., & Perrey, S.
(2013). Prefrontal cortex activity during motor tasks with additional mental load
requiring attentional demand: a near-infrared spectroscopy study. Neuroscience
Research, 76(3), 156-162.
Manes, F., Sahakian, B., Clark, L., Rogers, R., Antoun, N., Aitken, M., & Robbins,
T. (2002). Decision‐making processes following damage to the prefrontal cortex.
Brain, 125(3), 624-639.
Manoach, D. S., Schlaug, G., Siewert, B., Darby, D. G., Bly, B. M., Benfield, A.,
Edelman, R.R. & Warach, S. (1997). Prefrontal cortex fMRI signal changes are
correlated with working memory load. Neuroreport, 8(2), 545-549.
Marumo, K., Takizawa, R., Kawakubo, Y., Onitsuka, T., & Kasai, K. (2009).
Gender difference in right lateral prefrontal hemodynamic response while viewing
fearful faces: a multi-channel near-infrared spectroscopy study. Neuroscience
Research, 63(2), 89-94.
McCrae, R. R., & Costa Jr, P. T. (1997). Personality trait structure as a human
131
universal. American Psychologist, 52(5), 509.
Meshkati, N. & Lowewinthal, A. (1988). An eclectic and critical review of four
primary mental workload assessment methods: A guide for developing a
comprehensive model. In: Hancock, P.A., Meshkati, N. (Eds.), Human Mental
Workload. North-Holland, Amsterdam, 251-267.
Moffat, S. D., Hampson, E., & Hatzipantelis, M. (1998). Navigation in a “virtual”
maze: Sex differences and correlation with psychometric measures of spatial ability
in humans. Evolution and Human Behavior, 19(2), 73-87.
Monchi, O., Petrides, M., Petre, V., Worsley, K., & Dagher, A. (2001). Wisconsin
Card Sorting revisited: Distinct neural circuits participating in different stages of the
task identified by event-related functional magnetic resonance imaging. The Journal
of Neuroscience, 21, 7733– 7741.
Montello, D. R., & Pick, H. L. (1993). Integrating knowledge of vertically aligned
large-scale spaces. Environment and Behavior, 25(3), 457-484.
Moutafi, J., Furnham, A., & Crump, J. (2006). What facets of openness and
conscientiousness predict fluid intelligence score?. Learning and Individual
Differences, 16(1), 31-42.
Mulder, G. (1979). Mental load, mental effort and attention. In: Moray, N. Mental
Workload. New York, the USA, 299-325.
North, R.A., Riley, V.A. (1989). W/INDEX: A predictive model of operator
workload. Applications of Human Performance Models to System Design, 81-89.
Parikh, J., Neubauer, F. F., & Lank, A. G. (1994). Intuition: The new frontier of
management. Blackwell.
Parmenter, B. A., Shucard, J. L., Benedict, R. H., & Shucard, D. W. (2006).
132
Working memory deficits in multiple sclerosis: Comparison between the n-back task
and the Paced Auditory Serial Addition Test. Journal of the International
Neuropsychological Society, 12(05), 677-687.
O'Doherty, J., Kringelbach, M. L., Rolls, E. T., Hornak, J., & Andrews, C. (2001).
Abstract reward and punishment representations in the human orbitofrontal cortex.
Nature Neuroscience, 4(1), 95-102.
O’Donnell, R.D. & Eggemeier, F.T. (1986). Workload assessment methodology. In:
Boff, K.R., Kaufman, L., Thomas, J.P. (Eds.), Handbook of Perception and Human
Performance. Oxford, England, 1-49.
O’Laughlin, E. M., & Brubaker, B. S. (1998). Use of landmarks in cognitive
mapping:gen der differences in self report versus performance. Personality and
Individual Differences, 24(5), 595–601.
Overman, W., Graham, L., Redmond, A., Eubank, R., Boettcher, L., Samplawski,
O., & Walsh, K. (2006). Contemplation of moral dilemmas eliminates sex
differences on the Iowa gambling task. Behavioral Neuroscience,120(4), 817.
Ozcan, O.O., 2012. Exploring the effects of working memory capacity, attention,
and expertise on situation awareness in a flight simulation environment. MS.
Thesis, Middle East Technical University, Ankara, Turkey.
Persson, J., Herlitz, A., Engman, J., Morell, A., Sjölie, D., Wikström, J., &
Söderlund, H. (2013). Remembering our origin: gender differences in spatial
memory are reflected in gender differences in hippocampal lateralization.
Behavioural Brain Research, 256, 219-228.
Ragland, J.D., Turetsky, B.I., Gur, R.C., Gunning-Dixon, F., Turner, T., Schroeder,
L., Chan, R., & Gur, R.E. (2002).Working memory for complex figures: An fMRI
comparison of letter and fractal n-back tasks. Neuropsychology, 16, 370–379.
133
Reavis, R., & Overman, W. H. (2001). Adult sex differences on a decision-making
task previously shown to depend on the orbital prefrontal cortex. Behavioral
Neuroscience, 115(1), 196.
Reid, G.B. & Nygren, T.E. (1988). The subjective workload assessment technique:
A scaling procedure for measuring mental workload. Advances in Psychology, 52,
185-218.
Roscoe, A. H. (1992). Assessing pilot workload - why measure heart-rate, hrv and
respiration. Biological Psychology, 34(2-3), 259-287.
Roscoe, A. H. (1993). Heart-rate as a psychophysiological measure for in-flight
workload assessment. Ergonomics, 36(9), 1055-1062.
Rose, C. L., Murphy, L. B., Byard, L., & Nikzad, K. (2002). The role of the Big
Five personality factors in vigilance performance and workload. European Journal
of personality, 16(3), 185-200.
Sadalla, E. K., & Montello, D. R. (1989). Remembering changes in direction.
Environment and Behavior, 21(3), 346-363.
Schmidt, H., Jogia, J., Fast, K., Christodoulou, T., Haldane, M., Kumari, V., &
Frangou, S. (2009). No gender differences in brain activation during the N‐back
task: An fMRI study in healthy individuals. Human Brain Mapping, 30(11), 3609-
3615.
Simon, H. A. (1977). The logic of heuristic decision making. In Models of
Discovery (pp. 154-175). Springer Netherlands.
Smith, E. E., & Kosslyn, S. M. (2009). Cognitive Psychology. Mind and Brain.
Sohn, S. Y., & Jo, Y. K. (2003). A study on the student pilot's mental workload due
to personality types of both instructor and student. Ergonomics, 46(15), 1566-1577.
134
Solowij, N., Jones, K. A., Rozman, M. E., Davis, S. M., Ciarrochi, J., Heaven, P. C.,
Pesa, N., Lubman, D.I., & Yücel, M. (2012). Reflection impulsivity in adolescent
cannabis users: a comparison with alcohol-using and non-substance-using
adolescents. Psychopharmacology, 219(2), 575-586.
Speck, O., Ernst, T., Braun, J., Koch, C., Miller, E., & Chang, L. (2000). Gender
differences in the functional organization of the brain for working memory.
Neuroreport, 11(11), 2581-2585.
Stoltenberg, S. F., & Vandever, J. M. (2010). Gender moderates the association
between 5-HTTLPR and decision-making under ambiguity but not under risk.
Neuropharmacology, 58(2), 423-428.
Taylor, H. A., & Tversky, B. (1992). Descriptions and depictions of
environments. Memory & Cognition, 20(5), 483-496.
Turner, M. L., & Engle, R. W. (1989). Is working memory capacity task dependent?.
Journal of Memory and Language, 28(2), 127-154.
Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated
version of the operation span task. Behavior Research Methods, 37(3), 498-505.
van den Bos, R., Homberg, J., & de Visser, L. (2013). A critical review of sex
differences in decision-making tasks: focus on the Iowa Gambling Task.
Behavioural Brain Research, 238, 95-108.
Vicente, K. J., Thornton, D. C., & Moray, N. (1987). Spectral analysis of sinus
arrhythmia: A measure of mental effort. Human factors: The Journal of the Human
Factors and Ergonomics Society, 29(2), 171-182.
Vidulich, M.A. & Tsang, P.S. (1987). Absolute magnitude estimation and relative
judgement approaches to subjective workload assessment. Proceedings of the
Human Factors and Ergonomics Society Annual Meeting, 31(9), 1057-1061.
135
Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in
spatial abilities: a meta-analysis and consideration of critical variables.
Psychological Bulletin, 117(2), 250.
Waller, D., Knapp, D., & Hunt, E. (2001). Spatial representations of virtual mazes:
The role of visual fidelity and individual differences. Human Factors: The Journal of
the Human Factors and Ergonomics Society, 43(1), 147-158.
Wang, Y., & Ruhe, G. (2007). The cognitive process of decision making.
Watter, S., Geffen, G. M., & Geffen, L. B. (2001). The n-back as a dual-task: P300
morphology under divided attention. Psychophysiology, 38(06), 998-1003.
Wijers, A. A., Lange, J. J., Mulder, G., & Mulder, L. J. (1997). An ERP study of
visual spatial attention and letter target detection for isoluminant and nonisoluminant
stimuli. Psychophysiology, 34(5), 553-565.
Wierwille, W.W. & Casali, J.G. (1983). A validated rating scale for global mental
workload measure applications. Proceedings of the Human Factors and Ergonomics
Society Annual Meeting, 27(2), 129-133.
Willis, M. W., Ketter, T. A., Kimbrell, T. A., George, M. S., Herscovitch, P.,
Danielson, A. L., Benson, B.E. & Post, R. M. (2002). Age, sex and laterality effects
on cerebral glucose metabolism in healthy adults. Psychiatry Research:
Neuroimaging, 114(1), 23-37.
Wilson, G. F., Fullenkamp, P., & Davis, I. (1994). Evoked-potential, cardiac, blink,
and respiration measures of pilot workload in air-to-ground missions. Aviation
Space and Environmental Medicine, 65(2), 100-10
Wilson, G. F., & O'Donnell, R. D. (1988). Measure of operator workload with the
neuropsychological workload test battery. Advances in Psychology, 52, 63-100.
136
Wilson, G. F., & Russell, C. A. (2004). Psychophysiologically determined adaptive
aiding in a simulated UCAV task. Human performance, situation awareness, and
automation: Current Research and Trends, 200-204.
Yang, H., Zhou, Z., Liu, Y., Ruan, Z., Gong, H., Luo, Q., & Lu, Z. (2007). Gender
difference in hemodynamic responses of prefrontal area to emotional stress by near-
infrared spectroscopy. Behavioural Brain Research, 178(1), 172-176.
137
APPENDIX A
APPROVAL BY THE HUMAN SUBJECT ETHICS COMMITTEE
138
APPENDIX B
PARTICIPANTS’ PERSONAL DATA
Participant No Gender School Age
1 Female Hacettepe 23
2 Male Hacettepe 24
3 Male Hacettepe 21
4 Male Hacettepe 26
5 Male Hacettepe 26
6 Female Hacettepe 26
7 Male Hacettepe 26
8 Female METU 24
9 Male METU 24
10 Female METU 24
11 Male METU 25
12 Female METU 31
13 Male METU 31
14 Female METU 26
15 Male METU 26
16 Female Hacettepe 25
17 Male METU 26
18 Female Hacettepe 25
19 Male METU 25
20 Male METU 25
21 Female METU 24
22 Female Hacettepe 24
23 Female METU 23
24 Female METU 31
25 Male METU 28
26 Male Hacettepe 26
27 Female METU 23
28 Female METU 26
29 Female METU 23
30 Female METU 24
31 Female Hacettepe 25
32 Female Hacettepe 26
33 Female Hacettepe 26
34 Female Hacettepe 26
35 Female METU 31
36 Female METU 24
37 Female METU 25
139
APPENDIX C
PERSONALITY QUESTIONNAIRE
Türk Kültüründe Geliştirilmiş Temel Kişilik Özellikleri Ölçeği
Yönerge: Aşağıda size uyan ya da uymayan pek çok kişilik özelliği bulunmaktadır.
Bu özelliklerden her birinin sizin için ne kadar uygun olduğunu ilgili rakamı daire
içine alarak belirtiniz: (1) Hiç uygun değil, (2) Uygun değil, (3) Kararsızım, (4)
Uygun, (5) Çok uygun.
1 Aceleci 1 2 3 4 5 24 Pasif 1 2 3 4 5
2 Yapmacık 1 2 3 4 5 25 Disiplinli 1 2 3 4 5
3 Duyarlı 1 2 3 4 5 26 Açgözlü 1 2 3 4 5
4 Konuşkan 1 2 3 4 5 27 Sinirli 1 2 3 4 5
5 Kendine güvenen 1 2 3 4 5 28 Canayakın 1 2 3 4 5
6 Soğuk 1 2 3 4 5 29 Kızgın 1 2 3 4 5
7 Utangaç 1 2 3 4 5 30 Sabit fikirli 1 2 3 4 5
8 Paylaşımcı 1 2 3 4 5 31 Görgüsüz 1 2 3 4 5
9 Geniş / rahat 1 2 3 4 5 32 Durgun 1 2 3 4 5
10 Cesur 1 2 3 4 5 33 Kaygılı 1 2 3 4 5
11
Agresif(Saldırgan
) 1 2 3 4 5 34 Terbiyesiz 1 2 3 4 5
12 Çalışkan 1 2 3 4 5 35 Sabırsız 1 2 3 4 5
13 İçten pazarlıklı 1 2 3 4 5 36 Yaratıcı 1 2 3 4 5
14 Girişken 1 2 3 4 5 37 Kaprisli 1 2 3 4 5
15 İyi niyetli 1 2 3 4 5 38 İçine kapanık 1 2 3 4 5
16 İçten 1 2 3 4 5 39 Çekingen 1 2 3 4 5
17 Kendinden emin 1 2 3 4 5 40 Alıngan 1 2 3 4 5
18 Huysuz 1 2 3 4 5 41 Hoşgörülü 1 2 3 4 5
19 Yardımsever 1 2 3 4 5 42 Düzenli 1 2 3 4 5
20 Kabiliyetli 1 2 3 4 5 43 Titiz 1 2 3 4 5
21 Üşengeç 1 2 3 4 5 44 Tedbirli 1 2 3 4 5
22 Sorumsuz 1 2 3 4 5 45 Azimli 1 2 3 4 5
23 Sevecen 1 2 3 4 5
140
APPENDIX D
ENGLISH VERSION OF THE PERSONALITY QUESTIONNAIRE
Turkish Culture Based Basic Personality Traits Inventory
Instructions: There are many personality items below that fit or not fit you. Circle
the number for each item on how much you agree that it is appropriate for you on a
five point scale: (1) very inaccurate (2) inaccurate (3) unsure (4) accurate and (5)
very accurate.
1 Impetuous 1 2 3 4 5 24 Passive 1 2 3 4 5
2 Pretentious 1 2 3 4 5 25 Self-disciplined 1 2 3 4 5
3 Sensitive 1 2 3 4 5 26 Greedy 1 2 3 4 5
4 Talkative 1 2 3 4 5 27 Nervous 1 2 3 4 5
5 Self-assured 1 2 3 4 5 28 Genial 1 2 3 4 5
6 Cold 1 2 3 4 5 29 Angry 1 2 3 4 5
7 Shy 1 2 3 4 5 30 Hidebound 1 2 3 4 5
8 Sharer 1 2 3 4 5 31 Ill-mannered 1 2 3 4 5
9 Easy-going 1 2 3 4 5 32 Lethargic 1 2 3 4 5
10 Brave 1 2 3 4 5 33 Worried 1 2 3 4 5
11 Aggressive 1 2 3 4 5 34 Rude 1 2 3 4 5
12 Hard-working 1 2 3 4 5 35 Impatient 1 2 3 4 5
13 Backstabbing 1 2 3 4 5 36 Creative 1 2 3 4 5
14 Enterprising 1 2 3 4 5 37 Capricious 1 2 3 4 5
15 Well intentioned 1 2 3 4 5 38 Withdrawn 1 2 3 4 5
16 Sincere 1 2 3 4 5 39 Timid 1 2 3 4 5
17 Self-confident 1 2 3 4 5 40 Touchy 1 2 3 4 5
18 Temperamental 1 2 3 4 5 41 Tolerant 1 2 3 4 5
19 Philanthropic 1 2 3 4 5 42 Tidy 1 2 3 4 5
20 Capable 1 2 3 4 5 43 Fussy 1 2 3 4 5
21 Lazy 1 2 3 4 5 44 Prudent 1 2 3 4 5
22 Irresponsible 1 2 3 4 5 45 Determined 1 2 3 4 5
23 Compassionate 1 2 3 4 5
141
APPENDIX E
THE EVALUATION OF BASIC PERSONALITY TRAITS INVENTORY
R: Reversing Entry
Extraversion Conscientiousness Agreeableness NeuroticismOpenness To
ExperienceNegative Valence
4 12 3 1 5 2
R 6 R 21 8 11 9 13
R 7 R 22 15 18 10 26
14 25 16 27 17 30
R 24 42 19 29 20 31
R 32 43 23 33 36 34
R 38 44 28 35
R 39 45 41 37
40
142
APPENDIX F
INFORMED CONSENT FORM
143
APPENDIX G
MAZE COMPLETION TIME
Figure 23: Boxplot of Maze Completion Time
144
APPENDIX H
MAZE PATH LENGTH
Figure 24: Boxplot of Maze Path Length
145
APPENDIX I
IST PERFORMANCE
Figure 25: Boxplot of IST Performance
146
APPENDIX J
IST COMPLETION TIME
Figure 26: Boxplot of IST Completion Time
147
APPENDIX K
IST TRIAL NUMBERS
Figure 27: Boxplot of IST Trial Numbers
148
APPENDIX L
N-BACK PERFORMANCES
Figure 28: Boxplot of N-back Performances
149
APPENDIX M
FNIR MEASURES DURING MAZE TASK
Figure 29: Boxplot of Maze fNIR Measures
150
APPENDIX N
FNIR MEASURES DURING IST
Figure 30: Boxplot of IST fNIR Measures
151
APPENDIX O
FNIR MEASURES DURING N-BACK TASK
Figure 31: Bocplot of N-back fNIR Measures
152
APPENDIX P
THE PERSONALITY TRAIT STATISTICS
Table 60: Descriptive Statistics of Personality Traits
N Minimum Maximum Mean Std. Deviation
Extraversion 37 1.50 4.88 3.52 0.73
Conscientiousness 37 1.50 4.25 3.18 0.67
Agreeableness 37 2.63 5.00 4.14 0.53
Neuroticism 37 1.56 3.89 2.64 0.64
Openness 37 2.17 5.00 3.60 0.64
Negative valence 37 1.00 2.83 1.59 0.41
Valid N (Listwise) 37
Descriptive Statistics
153
APPENDIX R
THE PERSONALITY TRAIT RESULTS
Table 61: Detailed Results of the Personality Trait
Pr. No. PF1 PF2 PF3 PF4 PF5 PF6
1 4.5 1.75 4.88 3.11 4.5 1.33
2 2 2.13 3.5 2.78 3 1.83
3 3 3.13 4.13 2.78 3.33 1.5
4 2.75 3.38 3.88 2.11 3 1.17
5 4.13 1.5 3.63 2.22 4 2.83
6 3.75 3.25 4 2.78 3.17 2.17
7 3 2.25 4.25 1.89 4 1.5
8 1.5 2.38 4.75 3.11 2.67 1.5
9 3.13 4 3.88 1.56 4 2
10 3.25 3.75 4 2 3.17 1.5
11 4.13 3.25 2.63 3.89 3.17 1.5
12 3.63 4.13 2.75 3.33 2.67 1.33
13 3 3.13 4.5 2.44 4.33 1.33
14 3.13 3.38 4.13 2.56 3.5 1.67
15 4.25 3.5 5 1.89 4.17 1
16 4.63 3.88 4.63 2.56 4 1.17
17 4.63 3.75 4.88 3.22 5 1.67
18 3.63 3.13 4.38 2.11 4 1.17
19 3.38 2.38 4 3.89 3.17 2.5
20 3.88 4.25 4 2.89 3.67 1.5
21 3.25 3.75 4.13 2.56 2.17 2
22 3.13 3.88 4.13 3.22 3.17 1.83
23 3 2.25 4.75 1.78 3 1.17
24 3.75 2.88 3.63 2.67 3.67 1.67
25 4 3.38 4.25 1.67 4 1
26 4.25 3.75 4.25 2.33 4.5 1.83
27 3.75 3.25 3.5 3.56 4.67 2.33
28 3.5 3.63 4 3 3.5 1.33
29 3.5 3.63 4.25 2.22 4 1.33
30 3.25 3.38 4.25 3 3.17 1.33
31 3.75 3.75 4.13 3.11 4 2
32 4.88 3.25 5 1.78 3.67 1.33
33 2.75 3.38 4 2.33 3.33 1.17
34 3.5 2.63 3.88 1.67 3.33 1.5
35 2.38 3.38 4.63 3.78 2.5 1.5
36 3.75 2.25 4.38 3 4 1.5
37 4.5 3.13 4.25 3 4 1.67
154
APPENDIX S
THE PERSONALITY TRAIT HIGHER/LOWER RESULTS
Table 62: Higher/Lower Results of the Personality Trait
Pr. No PF1 PF2 PF3 PF4 PF5 PF6
1 higher lower higher higher higher lower
2 lower lower lower higher lower higher
3 lower lower lower higher lower lower
4 lower lower lower lower lower lower
5 higher lower lower lower higher higher
6 higher lower lower higher lower higher
7 lower lower higher lower higher lower
8 lower lower higher higher lower lower
9 lower higher lower lower higher higher
10 lower higher lower lower lower lower
11 higher lower lower higher lower lower
12 higher higher lower higher lower lower
13 lower lower higher lower higher lower
14 lower lower lower lower lower lower
15 higher higher higher lower higher lower
16 higher higher higher lower higher lower
17 higher higher higher higher higher lower
18 higher lower higher lower higher lower
19 lower lower lower higher lower higher
20 higher higher lower higher higher lower
21 lower higher lower lower lower higher
22 lower higher lower higher lower higher
23 lower lower higher lower lower lower
24 higher lower lower lower higher lower
25 higher lower higher lower higher lower
26 higher higher higher lower higher higher
27 higher lower lower higher higher higher
28 higher higher lower higher lower lower
29 higher higher higher lower higher lower
30 lower lower higher higher lower lower
31 higher higher lower higher higher higher
32 higher lower higher lower higher lower
33 lower lower lower lower lower lower
34 higher lower lower lower lower lower
35 lower lower higher higher lower lower
36 higher lower higher higher higher lower
37 higher lower higher higher higher lower
Recommended